An Intelligent and Secure Communication of AIoT enabled...

An Intelligent and Secure Communication of AIoT enabled Devices empowered with IPK Algorithm

a b b bMuhammad Adnan Khan , Muhammad Sarfraz , Muhammad Asif , Muhammad Saleem , bMuhammad Yousaf

a Department of Computer Science, Lahore Garrison University, Lahore, Pakistan.b School of Computer Science, National College of Business Administration & Economics, Lahore, Pakistan.

1. Introduction

AIoT is creating advancement which can irritate the present examples in Information and Communication Technologies (ICT). It has been proposed as a critical bit of related living. As it will be a basic piece of the lives of people there are a couple of challenges. The AIoT has made new characteristics by interfacing various devices to the structure, yet has similarly motivated security threat getting the opportunity to be basic issues as found in the continuous reports of illegal observation camera control and vehicle hacking, etc. IoT is as of now required to apply encryption to sensor contraptions in conditions with various constraints that have not previously been at risk to encryption. Improvements zone for cutting edge things,

Information Technology Communication and IPV6 (Internet show) are empowering quick arrangement of action of AIoT wherever on the world. It is assessed that billions of IoT contraptions will be sent in the next 5 years [1]. IoT approach is vast in number and used to given responses for an enormous number for enhanced issues. Regardless of the way that IoT has some portion of potential outcomes in the propelled world, in the midst of its course of action, it encounters a couple of issues with respect to (w.r.t) heterogeneity of devices, device character, contraption organization, a safety device to device correspondence (D-2-D, etc [2]. To e m p o w e r t h e r e c o n c i l i a t i o n a n d administration of heterogeneous IoT gadgets, models, for example, Ubiquitous Sensor Network (USN), Sensor Web Enablement

Abstract:

Artificial intelligence Internet of Things (AIoT) will be a necessary part of our lives in the near future. It will be found as quick cooperation in our surroundings through the related sensor-based system. To be sure, even in an indirect method, it will serve us in a couple of structures as esteem included organizations over the cell stages. With the AIoT structures that make usage of data, actually, the data collection from contraptions can in like manner be a goal of cyberattacks. Device to Device (D-2-D) interchanges in AIoT was planned alongside various shows, for instance, Constrained Access Protocol (CoAP). Its huge stresses in the course of action of AIoT are to ensure the security of mechanisms and D-2-D one place to another. Furthermore, present correspondence shows for AIoT are without reliability features. It is a result of this that countermeasures in perspective on encryption are starting at now getting importance. There is a requirement for a solid cryptosystem for D-2-D in AIoT. In this investigation paper, we present an encryption technique which is indicated as EPEB as a security answer for AIoT. The proposed methodology works with the message which shows special characters, numbers, and bits for data encryption and decryption. In authority, the end key isn't known so we would encryption to able have the option to gadgets data using particular keys and scramble packet per special key.Keywords: AIoT, D2D Communication, Encryption, Decryption, CoAP, secure communication algorithm.

LGU Research Jounral for Computer Sciences & IT 3(4) LGURJCSIT

Khan et al LGURJCSIT 2019

LGU Research Jounral forComputer Sciences & IT

Vol. 3 Issue 4, October - December 2019

1

ISSN: 2521-0122 (Online)ISSN: 2519-7991 (Print)

(SWE), and so forth, are proposed [2]. Here, the security of contraptions, (for instance, discount e x t o r t i o n , d a t a u p r i g h t n e s s ) , D - 2 - D correspondence, etc, are not tended to altogether. Cryptography is exhaustively named Symmetric, Asymmetric and Hybrid based [15]. Exactly when cryptography has a spot with the amiss sort, by then it has open and private keys. At present Public Key Cryptography (PKC) [6], [8], [10], [13], [16], [18] accept a key part in a couple of zones, for instance, Banking, Online purchasing, E-mail, etc., Due to this, there is the high peril of getting attacked [9], [19] through estimating the remarkable RSA riddle keys from general society type. A part of the progressing varieties of RSA with respect to their execution examination [3], [5], [11], [12], [14], [17], [19], [20], [22] and memory prerequisites of key [7].A segment of the PKC is appropriate for a multi-key age plot [20], [21], [23], [24] for capable sharing of information among the substances like IoT and Cloud enlisting. Here we have examined the multi-key-based cryptosystems with reuse of keys are according to the accompanying: Enhanced and Secured RSA based Key Generation (ESRKG) [4], Dual RSA [8], Trivial RSA [7], and N-prime RSA. In these varieties, the quality insignificantly depends upon the N-bit moduli and on account of this, the time-memory tradeoff in like manner gets extended. Regardless, the IoT based device has the inconsequential gear basic, for instance, low power and low estimation of around 2K bits. Remembering the true objective to achieve high-security quality, we propose here the IPKS plan for encryption for D-2-D correspondence in IoT.

I. Literature Review In [25], ABE is associated with assurance security for IoT in perspective on nonspecific Publish-Subscribe structure. By then we battle, IoT contraptions produce only two or three little bit of data and to perform encoding on two or three little parts of data both ABE and AES [42] encryption systems turn out the computational raised for IoT devices. A critical analysis of the security concerns of the internet of things (IoT) dissects the security issues and challenges and gives a well-characterized security system as the secrecy of the client's protection and security which could result in its more extensive selection by masses[22]. This Enabling information assurance

through PKI encryption in IoT m-Health gadgets introduces a framework dependent on Gateways (GW) that total wellbeing sensor information and resolve security issues through advanced testaments and PKI information encryption[23]. In this paper author looks at the improvement of a cloud-based, versatile IoT back-end structure and administrations dependent on top for managing and dealing with v e h i c u l a r d a t a i n v a r i o u s u s e c a s e circumstances: CAN data gathering, remote device blasting, Eco-driving, atmosphere projection, and guess. The fundamental variation is an Infrastructure-as-a-Service (IaaS) plan with a reference execution passed on an Open Nebula based cloud. The second cycle continues running on a private Platform-as-a-Service (PaaS) cloud-dependent on the Cloud Foundry arrange inside the premises of a vehicle supplier association. The two varieties have been viably evaluated and endorsed with benchmarks[24]. In the paper, the author centers around i n f o r m a t i o n i n g e s t i o n a n d a m a s s i n g perspectives, putting in verification issues and plans. The course of action proposed has been made and associated concerning the Sii-Mobility national splendid city adventure on f l ex ib i l i ty and t ranspor t jo ined wi th organizations. Sii-Mobility is grounded on Km4City theory and instruments for keen city-data accumulation and organization creation [25]. Various sequences of action keeping an eye on data aggregation while ensuring the security, for example, security, uprightness, approval, and openness, can be found in the composing [30]-[41]. Makers sketched out a middleware in light of the Pub-Sub plan in [27]. Here security of endorser's bit of leeway and private of disseminated substance are guaranteed by utilizing engaging Predicate Based Encryption (PBE) and CP-ABE. Basically in [26], [28], makers delineated an arrangement for Pub-Sub building using CP-ABE KP/plans . Here every supporter characterizes separating ventures as passage systems in light of this and in KP-ABE, crush performs leak of messages by executing encoded look on mixed qualities. Thusly, it propels message to proposed endorsers. To ensure message security Publisher scrambles message using CP-ABE and appropriates it. In [29], Tariq delineated security plans using IBE and ABE to engage privacy and approval. Here it

LGU Research Jounral for Computer Sciences & IT 3(4) LGURJCSIT 2

gadgets and different gadgets. In the third step, the send data will be encrypted with the help of the encryption key. In the fourth step, the encoded data will be placed in IoT gadgets. At whatever point any IoT gadget will get to the required data, the IoT gadget will affirm the other gadget and give the required data in the decrypted form. There are 9 major steps of proposed algorithms. 1-5 steps for sender device end and 5-9 steps for receiver device end.• Encrypt the message with an encryption key• Encrypt training bits with encryption key• Addition of encrypted message with encrypted training bits• Final encrypted message sends to receiver device using IoT platform• At receiver end separate encrypted message and encrypted training bits• Secure Communication of IoT based Devices using EPEB Algorithm.• Key generate from encrypted training bits• Decrypt encrypted message using key (those obtain from above step)• Finally, receive an original message at the receiver enddevice. In given method key will be created b/w IoT 1st device which send the information & other IoT device. It like a symmetric key, so the key remains the same at both the sender and receiver side.A. Encryption for Message at Sender Device: Once the key is selected then encrypt message and training bits with this key and adding an encrypted message with encrypted training bits and final encrypted message transferred to receiver device with want to communicate with sender device using IoT platform. The system of encryption by proposes approach can be effortlessly comprehended with help of piece outline figure 1.

empowers wholesalers to sign and encode events in the meantime by using IBE and in like manner engage productively controlling of mixed events (from merchants to supporters) by means of Searchable Encryption. Advance Subscribers affirm the imprints identified with all of the qualities (of an event) using CP/KP-ABE. Most of these plans delineated is sensible for nonspecific Pub-Sub models. Thusly a point by point considers is required for the credibility of altering these designs for IoT. From now on toward this way, we propose and execute improved designs for secure EPEB, which engages secure AIoT. It is the upgraded variety of Vigenere figuring. Security examination displays that the proposed upgraded outline is much secure as emerge from customary Vigene re-figure.

II. Proposed Intelligent Privacy Key (IPK) Algor i thms for AIoT Devices Communication In this research, an Intelligent Privacy Key(IPK) encryption algorithm is proposed for data security in AIoT gadgets correspondence in an intelligent way.

A. Highlights of proposed techniques: Some fundamental features of the proposed system for data protection are given beneath: Proposed method relies upon the symmetric key planning that is essentially speedier than an awry key calculation: • The proposed strategy relies upon the symmetric key calculation that is essentially speedier than an uneven key calculation. • Key create strategy is exceptionally unpredictable• what's more, strong. (inappropriate bullet)• The unique cryptographic key for each customer.• I t takes af ter a poly-alphabet ic substitution procedure that replaces plain substance character with various figure characters.• Repeat examination and cryptanalysis are incredibly problematic that makes our techniques much secure.

B. Proposed Method in IoT Framework: In the underlying advance, the IoT gadget will exchange the data at the IoT server. At the point when data will send that point in the second step, the key is acquired b/w IoT first



Table 1-9. Numeric, Alpha and special character values


Figure 1. Encryption Message Method

Figure 2. Message Encryption for Proposed IPK Algorithm

The formula for D2D encryption for final message is:ßf= þi+ Ðißf= Final encrypted Message text character in the proposed method. þi= Encrypted message in the proposed method. Ði= Encrypted training bits character in the proposed method.

Þ1 = ß 1 + €1 (ð %) = H + M (ð%) = 17 + 22 = 39 = $ {TABLE-1} þ2 = ß 2 + €2 (ð %) = + O (ð% ) = 36 + 23 = 59 ( 4 3 % ) = 1 6 = H { T A B L E - 2 } þ3 = ß 3 + €3 (ð %) = N + R (ð% ) = 21 + 42 = 63 (43%) = 20 = M{TABLE-3} . . . þ8 = ß8 + €8 (ð %) = + I (% ð) = 34+ 11 = 44 (%) = 2 = Z {TABLE-8}þ9 = ß9 + €9 (% ð) = L+ L (% ð) = 13 + 13 = 26 =

8 { T A B L E - 9 }þ10 = ß10 + €10 (% ð) = H+. (% ð) = 17 + 37 = 5 4 ( 4 3 % ) = 1 1 = B { T A B L E - 1 }þ11= ß11 + €11 (% ð) = R + C (% ð) = 0 + 11 = 1 1 = C { T A B L E - 2 }þ12= ß12 + €12 (% ð) = , + O (% ð) = 39 + 22 = 6 1 ( 4 3 % ) = 1 8 = K { T A B L E - 3 }þ13= ß13 + €13 (% ð) = P + M (% ð) = 22 + 19 = 4 1 = R { T A B L E - 4 }þ14= ß14 + €14 (% ð) = K + M (% ð) = 16 + 18 = 34= _ {TABLE-5}


1. Original Message Encryption with Key: (Wrong heading number?)

In our proposed method first letter set of message phrase & key is encoded by using table no. 1 and second letter set message phrase & key encoded by utilizing table no. 2 and vice versa. This is repeated again and again up to table no 9. At that point 10th letters in order of message phrase & key encoded by table no 1 vice versa. Encryption formula for message is:

Where,

þi =ßi + €iþi= Encrypted message in proposed method. ßi= Original message text character in the proposed method. €i= Key phrase character in the proposed method (for shorter length key repeat). ð = alphabet length in the proposed method.We can additionally simplify the proposed f o r m u l a a s :

þ1 = ß1 + €1 (% ð) {from table 1, in which Q=0,

R=1, S=2……. #=42} þ2 = ß2 + €2 (% ð) {from table 2, in which R=0, S=1, T=2……. Q=42} þ3 = ß3 + €3 (% ð) {from table 3, in which S=0, T=1, U=2……. R=42} . .. þ8 = ß8 + €8 (% ð) {from table 8, in which X=0, Y=1, Z=2……. W=42} þ9 = ß9 + €9 (% ð) {from table 9, in which Y=0, Z=1, A=2……. X=42}þ10 = ß10 + €10 (% ð) {from table 1, in which Q=0, R=1, S=2……. #=42}þ11 = ß11 + €11 (% ð) {from table 2, in which R=0, S=1, T=2……. Q=42} þ13 = ß13 + €13 (% ð) {from table 3, in which S = 0 , T = 1 , U = 2 … … . R = 4 2 }þ 14 = ß 14 + €14 (% ð) {from table 4, in which T=0, U=1, V=2……. S=42}Examples: Let us we consider this, our Message text is “H.NO#10, LHR, PK” & key phrase is “[email protected]” as shown in table 10.

Table 10. Message and Key Phrase

Table 11. Encrypted Message

2.Training Bits Encryption with Key:

The formula for training bits encryption is Ð i = µ i + € iÐi= Encrypted training bits character in the proposed method. µi= Training bits character in the proposed method. €i= Key phrase character in the proposed method (In the event that key length is shorter than the length of plain text then the key will be repeated). ð = alphabet length in the proposed method.Ð1 = µ1 + €1 (% ð) = N + M (% ð) = 23 + 22 = 45(43%) =2= S {TABLE-1} Ð2 = µ2 + €2 (% ð) = E + O (% ð) = 13 + 23 = 36 = {TABLE-2} . .

3. Final encrypted message

Now we add encrypted message and training bits by utilizing table 1-9 accordingly represented in table 13:

. Ð8 = µ8 + €8 (% ð) = + I (% ð) = 30 + 11 = 41 = V { T A B L E - 8 }Ð9 = µ9 + €9 (% ð) = N + L (% ð) = 15 + 13 = 28 = @ { T A B L E - 9 }Ð10 = µ10 + €10 (% ð) = E +. (% ð) = 14 + 37 = 5 1 ( 4 3 % ) = 8 = Y { T A B L E - 1 }Ð11 = µ11 + €11 (% ð) = T + C (% ð) = 2 + 11 = 1 3 = E { T A B L E - 2 }Ð12 = µ12 + €12 (% ð) = + O (% ð) = 35 + 22 = 5 7 ( 4 3 % ) = 1 4 = G { T A B L E - 3 }Ð13= µ13 + €13 (% ð) = P + M (% ð) = 22 + 19 = 4 1 = R { T A B L E - 4 }Ð14 = µ14 + €14 (% ð) = K + M (% ð) = 16 + 18 = 3 4 = _ { T A B L E - 5 }

Finally, your encrypted bits will be in shown in table 12:


Table 12. Encrypted bits

Figure 3. Decryption Message Method

In the proposed method at the receiver end, we received the message in the encrypted form like this “, BP216@AJJP9,4”. Now we separate encrypted training bits and messages using table 1-9 as shown in table 14.


Table 14. Encrypted bits and message

For the following above example: €1 = Ð1 - µ1 (% ð) = S - N (% ð) = 2 - 23 = -21(43%) = 22= M {TABLE-1} €2= Ð2 – µ2 (% ð) =- E (% ð) = 36 – 13 = 23 = O {TABLE-2} . . €8 = Ð8 – µ8 (% ð) = V - (% ð) = 41 - 30 = 11 = I {TABLE-8} € 9 = Ð 9 – µ 9 ( % ð ) = @ - N ( % ð ) = 2 8 - 1 5 = 1 3 = L { T A B L E - 9 } .€9 = Ð9 – µ9 (% ð) = @ - N (% ð) = 28 - 15 = 13 = L {TABLE-9} €10 = Ð10 – µ10 (% ð) = Y - E (% ð) = 8 - 14 = -6(43%) = 37= {TABLE-1}€ 1 1 = Ð 1 1 – µ 1 1 ( % ð ) = E - T ( % ð ) = 1 3 - 2 = 1 1 = C { TA B L E - 2 }€12 = Ð12 – µ12 (% ð) = G - (% ð) = 14 - 35 = -21(43%) = 22= O {TABLE-3}€ 1 3 = Ð 1 3 – µ 1 3 ( % ð ) = Q - P ( % ð ) = 4 1 - 2 2 = 1 9 = M { TA B L E - 4 }€ 1 4 = Ð 1 4 – µ 1 4 ( % ð ) = _ - K ( % ð ) = 3 4 - 1 6 = 1 8 = M { TA B L E - 5 }

At last, the proposed method generated decrypted key from training bits will be shown in table 15.


1st find encryption key from encrypted training bits, when the proposed system finds the optimum key then decrypts the given a message. The formula for D2D decryption for final e n c r y p t e d m e s s a g eßi =þi -€i

ßi= Final encrypted Message text character in the proposed method. þi= Encrypted message in the proposed method. €I= Encryption key character in the proposed method.

Figure 4. Key decryption from training bits for proposed IPK algorithm

Table 15. Encryption key generation.

T h e f o r m u l a o f o r i g i n a l m e s s a g e d e c r y p t i o n i s :

For the following above example: ß1 = Þ1 - €1 (% ð) = $ - M (% ð) = 39 - 22 = 17 = H {TABLE-1} ß2= Þ2 – €2 (% ð) = H - O (% ð) = 16 – 23 = -7 (43%) = 36 =. {TABLE-2} . ... ß8 = Þ8 – €8 (% ð) = Z - I (% ð) = 2 -11 = -9 (43%) = 34 = {TABLE-8} ß 9 = Þ 9 – € 9 ( % ð ) = 8 - L ( % ð ) = 2 6 - 1 3 = 1 3 = L { T A B L E - 9 }ß10 = Þ10 – €10 (% ð) = B - (% ð) = 11 - 37 = -26 (43% ) = 17 = H {TABLE-1}ß11= Þ11 – €11 (% ð) = C – C = (% ð) = 11 – 11 = 0 = R {TABLE-2}ß12= Þ12 – €12 (% ð) = K - O (% ð) = 18 – 22 = -4 (43% ) = 39 = , {TABLE-3}ß 1 3 = Þ 1 3 – € 1 3 ( % ð ) = R - M ( % ð ) = 4 1 – 1 9 = 2 2 = P { TA B L E - 4 }ß14= Þ14 – €14 (% ð) = _ - M (% ð) = 34 – 18 = 16 = K {TABLE-5}


Now proposed method use this key for d e c r y p t i n g t h e e n c r y p t e d m e s s a g e .

2. Decryption of Encrypted Message using Obtaining Key:

Figure 5 shown the decryption procedure given below:

Figure 5. Message decryption using the key for proposed IPK algorithm

Table 16. Decrypted message

Finally, we get the original message from the proposed method at the receiver end device.

Conclusion:

Strong Algorithms mechanism play a very strong role in different application domain like IoT, IoMT, AIoT, etc. In this article, proposed a new encryption method name IPK for secure communication for IoT devices. The proposed methodology worked with the message which shows special characters, numbers, and bits for data encryption and decryption. In authority, the end key isn't known so we would encryption to able have the option to gadgets data using particular keys and scramble packet per special key.

References

[1] B. S. Adiga, P. Balamuralidhar, M. A. Rajan, R. Shastry, and V. L. Shivraj, “An Identity Based Encryption Using Elliptic Curve C r y p t o g r a p h y f o r S e c u r e M2MCommunication,” in Proceedings of the first international Conference on Security of Internet of things, ser. SecurIT '12. ACM, 2012, pp. 68–74

[2] D . D ´ ı a z P a r d o d e Ve r a , ´ A . Sig¨uenzaIzquierdo, J. BernatVercher, and L. A. Hernandez Gomez, “AUbiquitous sensor network platform for integrating smart devices into the semantic sensor web,” vol. 14, no. 6.Multidisciplinary Digital Publishing Institute, 2014, pp. 10725–10752.

[3] ChandrasegarThirumalai, Senthilkumar M, SilambarasanR, Carlos Becker Westphall, “Analyzing the strength of Pell's RSA,” IJPT, Vol. 8 Issue 4, Dec. 2016 pp.21869-21874.

[4] Thangavel , M. , P. Varalakshmi, MukundMurrali, K. Nithya, “An Enhanced and Secured RSA Key Generation Scheme (ESRKGS),” in Journal of Information Security and application, Vol. 20, 2015, pp. 3-10.

[5] ChandrasegarThirumalai, Senthilkumar M, Vaishnavi B, “Physicians Medicament using Linear Public Key Crypto System,” in International Conference on Electrical, Electronics, and Optimization Techniques IEEEICEEOT, March 2016.

[6] Bellini, Emanuele, and Nadir Murru. 2015. “An Efficient and Secure RSA--like Cryptosystem Exploiting R'edei Rational Functions over Conics.” : 1–18.

[7] ChandrasegarThirumalai, “Review on the memory-efficient RSA variants,” I n t e r n a t i o n a l J o u r n a l o f P h a r m a c y andTechnology, Vol. 8 Issue 4, Dec. 2016, pp.4907-4916.

[8] Hung-min sun, Mu-en wu, Wei-chi ting, and M. JasonHinek “Dual RSA and its security analysis,” IEEE transactions on information theory, vol. 53, no. 8, August 2007.

[9] T Chandra Segar, R Vijayaragavan, “Pell's RSA key generation and its security analysis,” in Computing, Communications, and Networking Technologies (ICCCNT) 2013, pp. 1-5.

[10] Rives t RL, Shamir A, Adleman LA.1978, “Method for obtaining digital signatures and public-key cryptosystems”. Commun ACM.

[11] ChandrasegarThirumalai, “Physicians Drug encoding system using an Efficient and Secured Linear Public KeyCryptosystem (ESLPKC),” International Journal of pharmacy and technology, Vol. 8 Issue 3, Sep. 2016 pp. 16296-16303.

[12] Mayank Jhalani, Piyush Singh, Gaurav Shrivastava, “Enhancement over the variant of public-key cryptography algorithm,” in International journal of emerging technology and advanced engineering, Vol. 2, Issue 12, Dec. 2012.

[13] Chandramowliswaran, N., S. Srinivasan, and P. Muralikrishna. "Authenticated key distribution using a given set of primes for secret shar ing," Systems Science & Control Engineering 2015, Vol.3, Issue 1, pp. 106-112.

[14] Chandramowliswaran N, Srinivasan's, and ChandraSegar.T, “A Note on Linear based Set Associative CacheAddress System” International J. of Computer Science and Engg. (IJCSE) & India, Engineering Journals & 0975-3397, Vol. 4 No. 08 / pp. 1383-1386 / Aug. 2012.


[15] Forouzan BA.2007, “Cryptography and network security”. Special Indian Edition. Tata McGraw-Hill, p. 2011.

[16] ChandrasegarThirumalai, Senthilkumar M, “Secured E-Mail System using Base 128 Encoding Scheme,” International Journal of pharmacy and technology, Vol. 8 Issue 4, Dec. 2016 pp. 21797-21806.

[17] Ravi Shankar Dhakar, Amit Kumar Gupta, Prashant Sharma, 2012,” Modified RSA E n c r y p t i o n A l g o r i t h m ( M R E A ) ” advanceAdvancedComputing&Communication Technologies (ACCT).

[18] Chandramowliswaran N, Srinivasan.S and ChandraSegar.T, “A Novel Scheme for S e c u r e d A s s o c i a t i v e M a p p i n g ” T h e International J. of Computer Science and applications (TIJCSA) & India, TIJCSA Publishers & 2278-1080, Vol. 1, No 5 / pp. 1-7 / July 2012.

[19] He, Debian, et al. "Efficient and anonymous mobile user authentication protocol using self-certified public key cryptography for multi-server architectures." IEEE Transactions on Information Forensics and Security 11.9 (2016): 2052-2064.

[20] C h a n d r a s e g a r T h i r u m a l a i , SathishShanmugam, “Multi-key distribution scheme using Diophantine form for secure IoT communications,” IEEE IPACT 2017.

[21] Butun, Ismail, et al. "Cloud-centric multi-level authentication as a service for secure public safety device networks." IEEE Communications Magazine 54.4 (2016): 47-53.

[22] ChandrasegarThirumalai, Viswanathan P, “Diophantine based Asymmetric Cryptomata fo r c l oud Conf iden t i a l i t y and B l ind Signatureapplications,” JISA, Elsevier, 2017.

[23] Vasco, María Isabel González, Florian Hess, and Rainer Steinwandt. "Combined schemes for signature and encryption: The public-key and the identity-based setting." Information and Computation 247 (2016): 110.

[24] Shim, Kyung-Ah. "A Survey of Public-Key Cryptographic Primitives in Wireless Sensor Networks." IEEE Communications Surveys & Tutorials 18.1 (2016): 577-601.

[25] X. Wang, J.Zhang, E.Schooler, and M. Ion, “Performance evaluation of Attribute-Based Encryption: Toward data privacy in the IoT,” in Communications(ICC), 2014 IEEE International Conference on, June 2014, pp. 725–730.

[26] M. Ion, “Security of Publish/Subscribe Systems,” Ph.D. dissertation, University of Trento, Italy, May 2013.

[27] P. Pal, G. Lauer, J. Khoury, N. Hoff, and J. Loyall, “P3S: A Privacy-Preserving Publish-subscribe Middleware,” in Proceedings of the 13th International Middleware Conference, ser. Middleware '12, pp. 476– 495.

[28] M. Ion, G.Russello, and B.Crispo, “Supporting Publication and Subscription Confidentiality in Pub/Sub Networks,” in S e c u r i t y a n d P r i v a c y i n CommunicationNetworks, ser. Lecture Notes of the Institute for Computer Sciences, Social In fo rmat ics and Te lecommunica t ions Engineering, vol. 50, 2010, pp. 272–289.

[29] M. A. Tariq, “Non-functional Requirements in Publish/SubscribeSystems,” Ph.D.dissertation, Secure Communication of IoT-based Devices using EPEB Algorithm.

[30] Farooq, M. U., Waseem, M., Khairi, A., & Mazhar, S. (2015). A critical analysis of the security concerns of the internet of things (IoT). International Journal of Computer Applications, 111(7).

[31] Doukas, C., Maglogiannis, I., Koufi, V., Malamateniou, F., & Vassilacopoulos, G. (2012, November). Enabling data protection through PKI encryption in IoT m-Health devices. In 2012 IEEE 12th International Conference on Bioinformatics & Bioengineering (BIBE) (pp. 25-29). IEEE.


[32] Marosi, A. C., Lovas, R., Kisari, Á., & Simonyi, E. (2018, January). A novel IoT platform for the era of connected cars. In 2018 IEEE International Conference on Future IoT Technologies (Future IoT) (pp. 1-11). IEEE.

[33] Bellini, P., Nesi, P., Paolucci, M., & Zaza, I. (2018, March). Smart City architecture for data ingestion and analytics: Processes and solutions. In 2018 IEEE Fourth International Conference on Big Data Computing Service and Applications (BigDataService) (pp. 137-144). IEEE.


Preservation of Privacy of Big Data Using EfficientAnonymization Technique

Afia Naeem, Dr. Muhammad Rizwan, Dr. Fahad Ahmad

Department of Computer Science, Kinnaird College for Women, Lahore, Pakistan

1. Introduction

Big data can be termed as the study of big datasets generated from multiple sectors e.g. from systems, social media, etc. Data has become the raw material for production, a new source of economics and social value. Big data provides a valuable outcome to all the data analyst so that they could use it again. It has become an important topic for research purpose. Big data can help in the business purpose for the analysis purpose [1]. Data generated can be classified into two ways that are active data generation and passive data generation. In the active data generation, the owner of data is willingly providing data to the third party while in the passive the data owner is not aware that the data is being accessed by the third party.

Privacy is actually the protection of data from being exposed to the public network. The privacy of personal data is the most concerning aspect which should be achieved. Privacy

should focus on the usage of the data rather than a collection of data. Because if the data is not kept secured then it causes many threats. So for that, it should be modified according to the size of data as well as the unexpected use of it. To keep all the personal data and sensitive data private there are many anonymization techniques that help in the preservation of privacy [1]. The main goal of anonymization is actually to perform some masking operations on the data to protect the privacy of the individual with the insurance that it remains useful for researchers. The data contains the information related to the individual, organization, etc. The data contains three types of attributes which are as follows [2]:

Identifiers: These are the attributes that uniquely identify the individual e.g. Person name, CNIC number, etc.

Quasi Identifiers: Quasi Identifier (QI) are the attributes that are already known by everyone

Abstract:

Big data needs to be kept private because of the increase in the amount of data. Data is generated from social networks, organizations and various other ways, which is known as big data. Big data requires large storage as well as high computational power. At every stage, the data needs to be protected. Privacy preservation plays an important role in keeping sensitive information protected and private from any attack. Data anonymization is one of the techniques to anonymize data to keep it private and protected, which includes suppression, generalization, and bucketization. It keeps personal and private data anonymous from being known by others. But when it is implemented on big data, these techniques cause a great loss of information and also fail in defense of the privacy of big data. Moreover, for the scenario of big data, the anonymization should not only focus on hiding but also on other aspects. This paper aims to provide a technique that uses slicing, suppression, and functional encryption together to achieve better privacy of big data with data anonymization.

Keywords: Big Data, Anonymization, slicing, functional encryption, Privacy Preservation


Naeem et al LGURJCSIT 2019



14


and when they are taken together they can identify the individual e.g. Date of birth, Zip code, Gender.

Sensitive Attributes: These are the attributes that are not known to anyone and they contain sensitive information about the individual e.g. Salary of Person. They are unknown to the third party. The table below shows examples of these attributes.

Table I: Table of Attributes

Table I shows the example of different types of attributes in data. The Name is the Identifier, Gender and age are Quasi Identifier and Education is the sensitive attribute.

The information which is disclosed is of various types that are Identity disclosure, Membership disclosure, and Attr ibute Disclosure. The Identity disclosure is when a particular record that is released has linkage with an individual. Membership Disclosure is when the data which is to be published is extracted out from a large data which is sensitive. It is prevented from the third party from learning that the individual’s data is present in the data or not. Attribute disclosure is when the individual's new information is disclosed. The data which is r evea l ed makes i t e a sy t o know the characteristics of the individual. These types of information are disclosed when the data publisher publish the data, therefore all this information needs to be secured.

Big data include Vs (Velocity, Volume, Veracity, and Variety) that are equally important when analyzing the big data. The big data is based on the V-based characterization which has the main motivation of highlighting serious challenges i.e. storage, analysis, cleaning, etc [2]. Figure 1 shows the Vs of Big Data. The organization collects a variety of data from different data sources so that it is in large volume. This data is streamed at different speeds and they are of different kinds and formats. Storage of these datasets is challenging for the traditional database. So with this reason around

big data is stored and shared on the web which requires high security to keep that data private and protected from any harm or loss [3].

Fig. 1: Big Data V's

A. Ty p e s r e g a r d i n g I n f o r m a t i o n disclosure

There are basically two sorts of information disclosure that are Identity disclosure and Attribute disclosure. Identity disclosure is when an individual can be distinguished from the data being published. And Attribute disclosure is when the new information about any of the individual is released [4]. Identity disclosure allows the Attribute disclosure. K-anonymity is a technique that is meant to be a major backbone that helps in the protection of privacy to the data. These techniques also prevent Identity and Attribute disclosure.

B. Preservation of Privacy of Big Data

Big data benefits the industrial as well as research areas. The data nowadays is of different types and huge in volume. Big data have features such as structured, unstructured, semi-structured and heterogeneous [5]. Data generated from different social networks, the Internet, medical application and other sources is the big data that is huge and complex. This huge amount of data needs a way that can preserve the privacy of all the data [6]. Big data privacy is gaining concern as the amount of data is gradually increasing. With this huge amount of data, it is also important to preserve its privacy in a most efficient and reliable way [7]. The usage of data shall be focused rather than only


the collection of the data. Data should be protected before it is published. Unprotected data reveals sensitive attributes as well as identity information [1]. The major challenge is to protect the data from the third party which is in keen on collecting that data. Failure in keeping privacy leads to harm to the data and the individual 's personal information.

C. Anonymization of data

Anonymizat ion of da ta a ims a t maximizing the benefit by minimizing the individual risk. Anonymization is actually an approach in which the changes are performed on the data in such a way that the sensitive data and the identity of the individual are kept private and secure [8]. Different techniques are used for the anonymization of data that can hide the private data. It is the most important task to hide the identifier attribute. The identification of the key information is protected when anonymization is performed on data. Some details in the data are confidential which needs to be hidden from the third party and all the threats [4]. Data anonymization declares all the information that is used for queries as well as analysis while it maintains the sensitive data to be kept private Demand of Anonymization of data is because of the increasing occurrences of misuse of personal data and privacy issues [8]. Big data privacy is a domain that cannot be neglected at any stage.

The rest of the paper is arranged as follows. S e c t i o n I I d e s c r i b e s d i f f e r e n t anonymization techniques proposed previously. Moving on to Section III that is the Literature review. Section IV discuss the problem statement of the selected topic. The proposed solution is described in Section V. The result discussion is presented in Section VI. Finally, Section VII summarizes the paper.

2. ANONYMIZATION TECHNIQUES

For keeping the data privacy different anonymization techniques are carried out to anonymize the data. The following techniques are used for anonymization and the table below shows data:


Table II: Table Example Data

A. Suppression

Suppression is one of the anonymization technique in which an entire tuple or the attribute value is removed. The tupling is neglected. In the procedure basically, the original data is replaced by some special characters e.g. with (*) in the replacement of the data which is to be kept hidden [9]. The steric represents the data that is not supposed to be disclosed.This helps the data to be private and preserve from the third party to be known. The suppression technique on the sample data is presented as follows:

Table III: Table Suppression

B. Generalization

One other anonymization technique is known as Generalization. It can be said as Recoding. In this method, the value of an attribute is replaced by semantically consistent values, but that value is fewer specific ones [10]. Generalization replaces the exact value with some generalized one so that the details of that attribute are mot identifiable. The exact values are changed into a general range of data. The attribute value is generalized so with the help of this technique the third party cannot see the exact value which makes it complex to guess the exact value against something [9]. An example of generalization is as follow:

Table IV: Table Generalization

C. Bucketization

B u c k e t i z a t i o n i s s i m i l a r t o Generalization but the Quasi Identifiers or sensitive attributes are not modified. It divides

the records of data into partitions and each partition is given an id which is known as GID. The tuples in the partition have the same value of GID. After that Quasi Identifier Table and Sensitive Table are made. The data which is anonymized have set of buckets that have values of permuted sensitive attributes. The grouping which is made by bucketization is the same as grouping which is made in generalization. Bucketization has all original values of tuple, while the generalization has generalized values of tuple. The technique of bucketization work for the anonymization of data with high dimensions [9]. It provides much better utility, unlike generalization. But this technique don't stop the membership disclosure because values of Quasi Identifier are published in their original form so that is why it is easy to find the record of the individual. Sensitive attribute and Quasi Identifier need separation and it breaks down the correlation between them. The bucketization of the sample data given in Table 1 is as follows:

Table V: Table Quasi Identifier

Table V shows the Quasi Identifier where the Gender and Age are Quasi Identifiers and it is assigned GID. Now Table VI shows the Sensitive Attribute Table where Education is the sensitive attribute and they are also assigned GID.

Table VI. Table Sensitive Attribute

C. Slicing

This technique works on the flaws of the Generalization. Slicing works on data set in two different ways, i.e. vertical portioning of the dataset and horizontal portioning of the dataset [10]. In the vertical partitioning of the dataset, the dataset is grouped in a way that the attributes that are extremely correlated with each other are grouped into one column. Whereas in the horizontal partitioning the dataset is grouped in a way that the tuples are portioned into buckets. In each bucket, the value against each column is sorted for breaking the linkage between the different columns. Every bucket contains a subset of different tuples in it. Each tuple has multiple matching. Moreover, it provides better utility for the preservation of privacy. Slicing techniques slice the big data so that infrequent attributes break the collaboration among them [11]. This technique reduces the dimensionality of data. Slicing technique works efficiently on big data privacy preservation because it breaks association among the uncorrelated attributes. The slicing is implemented on the sample data taken. The following example shows the horizontal slicing and vertical slicing is given below:

Table VII: Table Slicing

These described techniques are kind of anonymization that work on the data for securing it [12]. In big data, the anonymization techniques become less effective. It needs to be more than the covering of data as well as generalizing it.

3. LITERATURE REVIEW

For the preservation of the privacy of big data different research has been done previously. In [12] Tiancheng Li, Ninghui Li, Senior Member, IEEE, Jian Zhang, Member, IEEE, and Ian Molloy introduced a method of Slicing for the Preservation of Privacy. The experiments in this paper show that slicing is an approach better than bucketization and it prevents the disclosure of membership. Later in [1] an efficient approach for privacy preservation of data


mining was done that included the use of combined techniques of randomization and anonymization [13]. The proposed technique in this paper protects the sensitive data with less information loss which prevent various types of attack. Later in 2015 Tomislav Krizan, Marko Brakus, Davorin Vukelic came up with in-situ anonymization of big data. A software d e s c r i p t i o n i s g i v e n f o r t h e i n - s u i t anonymization of big data which is distributed in c l u s t e r f o r m [ 1 4 ] . M o r e o v e r, m a j o r anonymization techniques were discussed in this paper that includes randomization, generalization. Further, in 2017 Abid Mehmood, Iynkaran Natgunanathan Yong Xiang, Song Guo, Senior Member, IEEE, Guang Hua, M e m b e r , I E E E h a v e d i s c u s s e d t h e anonymization technique that provided privacy to the data which include Generalization, Suppression, Anatomization, Permutation, Perturbation. Later in [3] Nivedita Elanshekhar and Rajashree Shedge gave a solution that uses the Suppression Slicing method. In this paper, this approach was performed on the attributes which have similar values for better privacy and for better utility. The method follows the procedure that the data goes under MapReduce method and then suppression slicing is implemented. Furthermore, in [15] Brijesh B. Mehta and Udai Pratap Rao used the scalable k-anonymization approach using the MapReduce technique for the privacy preservation of big data. In this paper, an algorithm named as k anonymizat ion using MapReduce was introduced for the privacy preservation of big data publishing. This approach was compared with the existing approaches. In 2018 an approach was proposed which was a scalable approach for big data multidimensional anonymization which was based on the MapReduce.The idea was actually to partition the data set into small data sets using the MapReduce method. Then in [16] the same year 2018 an enhanced privacy preservation auction scheme was used that included an additional verification mechanism.

4. Problem Statement

Conserving the anonymization of big data sets is an important deal to be dealt with. When the data is anonymized, it means that all personal data is eliminated. In the era of big data, the data anonymization techniques tend to fail

because of the thousands of data points for the individual. These simple techniques don't fully perform efficiently like up to the mark in preventing from disclosure of identity. Suppression is easy to implement but the data quality is reduced drastically. Generalization fails when it comes to high dimensional data because of multiple dimensions and the data tends to lose much information [7]. Moreover, big data have a linkage of information that needs to be removed. Data anonymization needs to be much more than just masking the data and generalization of it. The existing models are lacking in managing large datasets. These techniques alone cannot preserve the privacy of the big data, Data anonymization techniques need to be improved in a way that it focuses on the 3Vs of big data i.e. Volume, Velocity, and Variety [1]. It should become more efficient so they can make a positive effect on keeping the big data protected from any loss or attack. There is a need for a new technique that carefully analyzes that the anonymized data is exposed to any harm or not.

5. Proposed Solution

Increasing the growth of data needs a solution that can provide preservation of the privacy of this big data. Many solutions have been implemented to keep the data private but when it comes to big data privacy those were unable to perform better [14]. The proposed solution in this paper carries three techniques which all together can work in a way that can provide the data privacy to be kept preserve from the third party and hackers. Figure II below shows the steps that are involved in the proposed solution followed by the explanation of these steps:

Fig. 2: Proposed Solution


A. Implementing Slicing on Data

Firstly, the slicing is implemented on the big data where the data is portioned both horizontally and vertically with the preceding procedure as it is discussed before in Section II. Implementing this technique have many advantages that assist in the preservation of the privacy of big data. The technique of slicing provides big data with privacy as it preserves the data utility as compared to the techniques of generalization and also provides the disclosure of attributes. Each tuple in the slicing has more than one matching within so that is why the slicing helped in keeping the privacy of big data [10]. With the help of the slicing of data, the data dimensionality is reduced. The correlation between the attributes is kept maintained. Moreover, it groups the attributes that have much correlation between them.

B. Implement Suppression

Secondly, after the data is sliced both vertically and horizontally the sliced data now undergo suppression where the data is kept hidden by replacing the sensitive attribute or sensitive tuple with (*). This helps the confidential attributes to be hidden from the hackers. It will make the data complex to be known. The hackers have to apply guesses in order to know the original data which is hidden under the steric.

C. Implement Functional Encryption

a) Encryption: Encryption is actually a process that involves encoding some information or any message in a way that the information is only accessed by the authorized parties and no other can access it. It is simply to mystify the information so that it is hidden from the unauthorized parties which have complete or any partial access over the information[19]. The message which has to undergo the encryption is called the plain text and when it is encrypted it is called ciphertext. The ciphertext can only be changed into the plaintext by using the decryption process [15]. For the purpose of encryption and decryption, it makes use of a key that is known only to the sender and the receiver. Encryption helps the data to be hidden from the third party or any attack. The encryption scheme makes use of the encryption key which is generated with an algorithm [20] [21] [22]. The

key is shared with both the communication parties at both the end i.e receiver end and sender end. For the encryption of data, many different encryption techniques are used. But when it comes to the encryption of big data, which is in huge volume the traditional encryption schemes fail to perform efficiently on it. a) Func t iona l Encrypt ion : Functional Encryption is a type of encryption that works on big data. Functional Encryption is a type of encryption that helps to keep the privacy of the big data preserve. The traditional encryption techniques only help to encrypt the limited amount of data [17]. When it comes to encrypting the big data, those techniques don't perform up to the mark. But the functional encryption helps to perform better when it comes to big data. Functional Encryption is a type of public-key cryptography It doesn't encrypt any other function than that specific one. This type of encryption differs from the traditional ones in a way that the generated ciphertext of the provided plain text is only decrypted by a specific recipient. While in the functional encryption the group of people can encrypt the message without knowing anyone [18]. Moreover, the traditional encryption encrypted all the data or nothing while in this selected data can be encrypted. The pictorial representation of functional encryption is shown in Figure III. The technique of functional encryption lets the user to only know any specific functionality without knowing what the rest of the data contains [17].

Fig. 3: Functional Encryption


In figure III the encrypted data is sent from the data owner to the server. The data user passes the function F to the trusted authority which in return sends the function token to the data owner. The function token is represented by the TF. The token function is sent from the data owner to the server. As a result, the server filters out important things or needed information. The filter helps in a way that the whole data is not known to the data user. Only a part of a function is known to the data user. The server after filtering send the function result to the data used so that the person can use that data.

The following box shows the proposed solution algorithm:

These steps of algorithm are carried out step by step on the big data so that it can be preserved and kept private. First, the data is received in slicing function where it is sliced both vertically and horizontally. Then the sliced data is received to the suppression function where the data is suppressed with the steric. In the end, it goes to functional encryption function whether a particular function or whole data is encrypted according to the demand of the user. In functional encryption function, the pair of master key and public key is generated.

Then a secret key (sk) for the value of k with the help of the master secret key using the key generator is generated. After that encrypts the message (m) with the help of the public key. Then the secret key (sk) is used to compute the function F (k, m) from c. The secret key holds the specification of only decrypt specific functionality and restricts the decrypting of other data. Encryption and Decryption functions are used to take the input and perform encryption and decryption. The following Figure IV shows the functional encryption in summarized form.

Fig. 4: Functional Encryption Summarized

The functional encryption includes four algorithms that include Setup, Encryption, Decryption, and Keygen. The Setup algorithm output the master key (mk). In the Keygen algorithm, the master key is taken as input and also some of the description related to function. Taking them as input the algorithm outputs a key that is only specific to the function which is denoted by the sk[f]. After the data undergo suppression by masking the tuples or attributes with a steric the data is passed through this functional encryption where the big data is encrypted and only a part of ciphertext is revealed to the user. Users cannot know nothing except that portion that is revealed. This helps in keeping the big data privacy to be preserve. With this feature of functional encryption along with anonymization techniques help the data privacy to preserve much better.

This proposed method which is discussed above can help in preventing hackers to hack the data and also from the third party to keep the big data anonymous. Moreover, it can work on big data which the other techniques fail to perform up to mark. Previously only the anonymization techniques were implemented for privacy reasons. But this solution can provide improved and better results and also in an efficient way.


6. RESULT & DISCUSSION

T h e p r o p o s e d s o l u t i o n , w h e n implemented on the big data, can result in a better privacy preservation. The data first is portioned in the vertical form or horizontal form. It assembles the attributes having a correlation with each other. Then the data is suppressed by hiding confidential attributes or tuples. After that, the functional encryption is applied that encrypts the function that is to be kept private with the help of the key. All the data is not encrypted but only a specific portion is hidden from the hackers. This all procedure result is privacy that is complex to be harmed and in each step of the big data lifecycle, it can preserve privacy. This proposed solution can perform better than the traditional anonymization alone can perform and can be automated with the increasing 3 V's (volume, velocity, and volume) in the big data.

7. CONCLUSION

Big data is growing every day and it contains private information which needs to be preserved. It is a term that is used for the data which is complex and of huge volume. It is used in the analysis and for decision purposes. At every stage of the big data life cycle, it requires high privacy. So, privacy is a major concern that needs to be focused. Privacy is a factor that needs not to be compromised because the disclosure of private information leads to harm of data and threats. For the privacy of this big data, the anonymization techniques were being used that include generalization suppression, etc. The technique of generalization or suppression alone was not able to handle when it comes to big data that is also the high dimensional data. One of the anonymization techniques is the slicing which is best for big data as it provides the portioning. Applying suppression and functional encryption it can give results up to the mark. If functional encryption is implemented properly then it can reduce the privacy challenges that big data may face. The preservation of the privacy of big data is a major emerging field that cannot be ignored because with the gradual increase in data it is becoming more complex.

REFERENCES

[1] A. Mehmood, I. Natgunanathan, Y. Xiang, G. Hua, and S. Guo, "Protection of Big Data Privacy," IEEE Access, vol. 4, pp. 1821-1834, 2016.

[2] M. Dave and J. Kamal, "Identifying big data dimensions and structure," in 2017 4th International Conference on Signal Processing, Computing and Control (ISPCC), Solan, India, 2018.

[3] N. Elanshekhar and R. Shedge, "An effective anonymization technique of big data using suppression slicing method," in 2017 In te rna t iona l Confe rence on Energy, Communication, Data Analytics and Soft Computing (ICECDS), Chennai, India, 2017.

[4] P. C. Kaur, T. Ghorpade and V. Mane, "Ana lys i s o f da t a s ecu r i t y by u s ing anonymization techniques," in 2016 6th International Conference - Cloud System and Big Data Engineering (Confluence), Noida, India, 2016.

[5] M. D. A. Praveena and B. Bharathi, "A survey paper on big data analytics," in 2017 International Conference on Information Communication and Embedded Systems (ICICES), Chennai, India, 2017.

[6] S. Singh and N. Singh, "Big Data analytics," in 2012 International Conference on Communication, Information & Computing Technology (ICCICT), Mumbai, India, 2012.

[7] Q. Tan and F. Pivot, "Big Data Privacy: Changing Perception of Privacy," in 2015 IEEE I n t e r n a t i o n a l C o n f e r e n c e o n S m a r t City/SocialCom/SustainCom (SmartCity), Chengdu, China, 2015.

[8] A. Kumar, M. Gyanchandani and P. Jain, "A comparative review of privacy preservation techniques in data publishing," in 2018 2nd International Conference on Inventive Systems and Control (ICISC), Coimbatore, India, 2018.

[9] T. Karle and D. Vora, "PRIVACY preservation in big data using anonymization techniques," in 2017 International Conference on Data Management, Analytics and Innovation (ICDMAI), Pune, India, 2017.


[10] P. Goswami and S. Madan, "Privacy p re se rv ing da t a pub l i sh ing and da t a anonymization approaches: A review," in 2017 International Conference on Computing, Communication and Automation (ICCCA), Greater Noida, India, 2017.

[11] P. C. Kaur, T. Ghorpade and V. Mane, "Ana lys i s o f da t a s ecu r i t y by u s ing anonymization techniques," in 2016 6th International Conference - Cloud System and Big Data Engineering (Confluence), Noida, India, 2016.

[12] T. Li, N. Li, J. Zhang and I. Molloy, "Slicing: A New Approach for Privacy Preserving Data Publishing," IEEE Transactions on Knowledge and Data Engineering , vol. Volume 24, no. Issue 3, pp. 561 - 574, 2010.

[13] M. Sharma, A. Chaudhary, M. Mathuria, S. Chaudhary and S. Kumar, "An efficient approach for privacy preserving in data mining," in 2014 International Conference on Signal Propagation and Computer Technology (ICSPCT 2014), Ajmer, India, 2014.

[14] T. Križan, M. Brakus and D. Vukelić, "In-situ anonymization of big data," in 2015 38th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), Opatija, Croatia, 2015.

[15] B. a. U.P.Rao, "Privacy preserving big data publishing: a scalablek-anonymization approach using MapReduce," IET Software, vol. 11, no. 5, pp. 271-276, 2017.

[16] W. F. W. a. C. W.Gao, "Privacy-Preserving Auction for Big Data Trading Using H o mo mo r p h i c E n c r y p t i o n , " i n I E E E Transactions on Network Science and Engineering, 2018.

[17] K.Takashima, "Recent Topics on Practical Functional Encryption," in Second International Symposium on Computing and Networking, Shizuoka,Japan, 2014.

[18] A. Boneh, "Functional Encryption:A New Vision for Public Cryptography," vol. 55, pp. 56-64.

[19] P. S. a. D. K. Kaur, "Database Security Using Encryption," in 015 International C o n f e r e n c e o n F u t u r i s t i c Tr e n d s o n Computational Analysis and Knowledge Management (ABLAZE), Noida, India, 2015.

[20] C. Matt and U. Maurer, "A Definitional Framework for Functional Encryption," in 2015 IEEE 28th Computer Security Foundations Symposium, Verona, Italy, 2015.


Cloud Storage Security Using Blockchain Technology Areeba Rahman, Dr. Muhammad Rizwan, Dr. Fahad Ahmad

Department of Computer Science, Kinnaird College for Women, Lahore, Pakistan

1. Introduction

Blockchain-based research had been done on the safety and security of bank money between peer groups and any third party. Blockchain is a ledger for transactions and saves from hacking. Services for remote cloud storage have increased over the past few years. The real problem is transferring data into an outside environment that no one can access that specific data other than the owner. Many ways can be found to secure data. Services that gives storage for data, to access that data and backups the data are easy to use. They also make life easy but here is a problem of trusting the third party. We handover our data to the third party. To overcome this problem, one way is to encrypt the data or any record. Cloud security provides this defensive approach. But encryption is difficult to handle. Encrypted data becomes secure In the blockchain, all the transactions are kept encrypted according to the rule that is defined in its software. Bitcoin that is electronic money uses blockchain knowledge. It provides transparency to the whole network. Usually, data

is stored in core databases that are less secured as compared to the blockchain because it gives more security and safety of data. Even if the database gets damaged because of attacks, it can be overcome with blockchain. Because of these facts and figures, this technology can be implemented not only in bitcoin but also in cloud computing, the Internet of things (IoT), healthcare and much more. The healthcare industry is incorporating IoT based solutions swiftly. Basically, blockchain is the future of cloud storage. It has been applied in many IT atmospheres also because of the efficiency and availability of cloud computing. Blockchain is a decentralized data structure. Many companies and indus t r ies provide the i r s towage substructure and cloud storage in a decentralized manner. They use servers on their own hosts in their offices that are quite costly and expensive. It is not easy to manage in-house servers because of so many facts one is their cost and management. But a convenient solution is Amazon S3. Services like amazon are totally d i f f e r e n t . I t i s u n b e l i e v a b l e . T h e s e accessibilities keep us safe from the downsides. To secure our very personal, private and

Abstract:

Data is increasing with increasing Internet technology. To handle the large data, more applications choose to enlarge storage capacity via Cloud plate form. It will not a surprise if we say most organizations have moved towards the cloud. While using the cloud, we have to keep our trust for our sensitive and private data in third parties and the data is usually not encrypted. But we need to implement nearly procedures for the assurance of our reserved data. This will be occupied by blockchains. Blockchain has been a center of attention as a next-generation goal because of its security. A comprehensive approach is used in this paper by signifying diverse blockchain methods to protect cloud computing.

Keywords: cloud storage, Trusted Third parties, Blockchain


Rehman et al LGURJCSIT 2019



23


sensitive data we have to keep our trust in these third-party tools. In other words, we are dependent on them just to secure our data. It can be stolen or hacked as well. In blockchain technology, the data is encrypted first and divided into fragments and then distributed among distributed nodes in many countries. Blockchain provides these incredible features that are not possible before.Ø C o m p l e t e r e d u n d a n c y a n d t r u e decentralizationØ Complete privacyØ Cost Deduction

2. THE BLOCKCHAIN DEFINED

The blockchain is considered the next big revolutionizing technology after the Internet, as it reinvents our way of working and living. In 2008, a scholar who applied the numerical cryptocurrency known as Bitcoin primary presented the impression of a blockchain. The blockchain is fundamentally an important slice of the process of Bitcoin. Since then there have been many cryptocurrencies with very advanced features, such as Ethereum, which introduces intelligent contracts. The main features of the blockchain are shown in Figure 1. From the exchange of information to money transfer and other belongings that require online transactions, everything involves a reliable intermediate. This trusted mid-way is accountable and takes all the responsibility in case of any failure and handles all the glitches that are related to security. The need for a central authority between different companies or parties to carry out multiple functions like data transaction and financial processing has been eliminated by blockchain technology via using straight, undisputable and distributed open accounts [1].

The network users use the public ledger which is a distributed and shared database. The public ledger is a kind of record that cannot be in ter fered wi th and very secured via cryptographic key distribution and it keeps the records of all of the transactions that have been done among the network users. The property that makes blockchain technology permanent, unchallengeable and irretrievable is that users can view the transactions which are related to them any time they want, but there is a process of validation and once the data or transaction is validated and authenticated, then it can't be deleted nor modified and revised. There is already a defined criterion for the network users to verify their transaction without confirmation and validation of any significant authority which includes validation, confirmation, agreement, and consensus. It has a great impact on the charge because it reduces the cost. Moreover, it also reduces the chances of data loss that usually happen because of a single point of failure and synchronization have already been done among all of the network users. Therefore, blockchain even assures privacy and security along with other noticeable features in which authentication, validation, decentralization, transparency and much more [2]. The evolution that came along with blockchain technology is high security. It also provides an innovative idea of software-defined parameters. The main thought behind this idea is, before starting the communication, it creates a secure and strong channel first. This channel is associated with a centralized controller. The idea of the software-defined parameter is getting a lot of attention and consideration. A problem that provoked from the dependency on the third-party authorization which resulted in a single point of failure has also been solved by blockchain. It permits all the associates and network users to maintain a ledger. This ledger contains all of the transaction data and other material. Other than maintaining the ledger, blockchain also allows the users to update the ledger so that the correctness and integrity can be maintained whenever there is new data or a transaction is made. Recently used research spaces such as cloud, Internet of Things (IoT), edge computing, cloud computing and much more, are based on those entities that are centralized. Whereas, blockchain provides the opportunity to eliminate these centralized controller entities if


Figure 1. Blockchain Characteristics

these research areas directly apply the blockchain technology. Therefore, the blockchain will benefit several developing technologies, comprising smart cities, banking, and the Internet of Vehicles The blockchain has broker-free (P2P-based) features. P2P means peer to peer or person to person transactions. The needless fee can be eliminated without having the permission of the third party. Therefore, with blockchain, many people can own the transaction information which makes it difficult to hack. Ultimately, security expenditures are saved. All of the transactions are approved spontaneously and logged by maintaining a record. Hence, reliability, swiftness, and promptness are guaranteed. In this way, transparency can be increased because of the open access and source to the transactions. It can also reduce supervising cost. It also provides a feasibility to the system so that it can be simply applied, linked and Furthermore, the scheme can be effortlessly applied, connected, and long-drawn-out. There are numerous continuing lessons to reinforce safety by means of these features of blockchain. The utmost significant slice of the blockchain is safety and security connected to the private key cast-off in encryption and here is training on how to defend the private key. An assailant tries a “reuse attack” and additional bouts to get the private key stowed in an aristocrat's scheme in command to menial the bitcoin. The assailant can hack the bitcoin meanwhile the data may have seeped if the assailant can get the private key. To resolve this difficulty, pieces of training on smearing together hardware and software safeties for favorable connections are continuing. In adding, Blockchain guarantees no dual expenditure occurs. Transactions are comprised, that is, here are no two transactions that apply the similar even of coins. This is understood by business authentication work in Blockchain. Some approaches and models have been discussed in this paper to secure cloud storage with different blockchain models.

3. BLOCKCHAIN SECURED CLOUD STORAGE USING CHAINFS

This exertion boons ChainFS, a bridge scheme that safeguards cloud storing facilities by means of a slightly reliable Blockchain.

ChainFS toughens the cloud storing safety in contradiction of splitting rounds . The chains Bridge delivers the end workers through a file scheme border. Within, ChainFS supplies information records in the cloud and spreads to the blockchain negligible and essential functioning for key delivery and cataloging of file operations. We organize and carefully assimilate the ChainFS scheme on Ethereum and S3FS with FUSE patrons and Amazon S3 cloud storing. We amount the presentation of the scheme and display little overhead.

A. S E C U R I T Y A N A LY S I S W I T H CHAINFS

ChainFS influences the Blockchain to add forking-attack pliability to an encoded file scheme on the cloud. For equal open key almanac and file scheme processes, it upholds a lined record of application-dependent entrances and charts the record to Blockchain. Forcing attacks in the open almanac means that the cloud has dissimilar key terms (of the same person) for dissimilar customers. For the customer to receive the requisite, recall that it forms the presence of the compulsory in the almanac snap with abridgment as a checked log entrance in the blockchain. Two record entrances must be checked in the blockchain at the similar time to let two patrons to receive two cleft bands. . For the SUNDR attack (SUNDR is a network file system designed to safely store data on untrusted servers) the stowage server can extant two encoded and dissimilar files to two diverse customers. These two measures are logged in the native logs of these two customers. In adding, the cloud server records its own global log version in the Blockchain. Native records will be likened to the worldwide record


Figure 2. System Overview

during log certification and auditing to exam whether the native record is a subsection of the worldwide record and if there is a fissure of the steadiness of the storing. In order to avoid the attack, Blockchain must be forced to fork itself. It is difficult in a large community blockchain to divide chunks that are established (e.g. after 6 times).

B. PERFORMANCE

We take three kinds of machineries in this experimentation system. Mainly, we create an Amazon S3 AWS account and route cloud cases. Furthermore, we route our FUSE patrons nearby. The customer mechanism has an Intel(R) Xeon(R) CPUE5-2680 v3 CPU with a memory of 2.50GHz and 10 GB Then, the Blockchain is a track on three server engines with the subsequent requirement: 2.70GHz and 8 MB hoard Intel8-core i7-6820HK microchip, 32 GB Ram and 1 TB Disk. File Create/Write Performance: In tests, we primary usage LFS minor file standards to produce 1000 small files of 1 KB to 100 KB sizes. We practice the Linux dd value to produce files and ration phase. Regular time and standard deviation are stated. In this situation, the files are produced by means of arbitrary content so that the abridgments to be placed on Blockchain alteration and ChainFS has no partial advantage. The consequences of the minor file experiment are shown in figure 2. Our ChainFS has up to 35 percent overhead presentation (with 10 KB files) to parallel the best situation that turns an S3FS without a Blockchain. As the files produce large, the overhead shrinkages. The routine becomes unbalanced, particularly when files are too minor. The participation of Blockchain does not upsurge much standard deviation, and we are unsure that this is for the reason that the actual cloud connection is rather undefined. We also convey out experimentations with big files following a related process. Files with a file mass of amid 1 MB and 1 GB are created. We degree the period of implementation and state metrics in a similar method as the minor file situation. The outcome is shown in Figure 3. The overhead of the Blockchain upsurges as the file produces greater and spreads a supreme of 28 percent (1 GB). In the outsized file arrangement, the blockage scheme handovers data above the Internet .

File Read Performance: We perform experimentat ion to est imate ChainFS' evaluation dormancy. The customer engine primary turns a script in the testing to generate 100 mutable scope records (from 1 KB to 100 KB). It formerly jolts a sequence of Linux CAT instructions to read the files above and over again. We portion the period consumed in the next phase (i.e. CAT commands) in this testing. On usual, ChainFS enhances about 30 percent overhead to the fixed cloud file arrangement read pathway. As the file produces huge, the overhead stays continuous .

4. BLOCKCHAIN-BASED ACCESS CONTROL SCHEME FOR CLOUD STORAGE

In this approach, a multiple worker system prototypes for controlling access is used for data sets to be stored in a cloud atmosphere that is not trusted. Just like any not trusted environment, cloud storage requires the ability to secure information sharing. This approach allows access to data/information which is stored in the secured cloud without the participation of the provider. The main tool which is used to access the control mechanism is a text-policy encryption scheme based on static attributes with dynamic attributes. The proposed scheme delivers an immutable record of all meaningful data security events, such as key generation, access policy assignment, access request, change or revocation, using a blockchain-based decentralized ledger. We suggest a set of cryptographic procedures to ensure the privacy of secret or private key cryptographic operations. Only hash code ciphertexts are transmitted via the blockchain ledger. Our system's prototype is implemented with intelligent contracts and tested on the Ethereum blockchain platform . The aim of the problem-solving approach is to develop an access control model based on blockchain transactions, data storage in untrusted storage and the implementation of Ethereum smart contracts based on attribute-based encryption. We use a model of access control based on attributes. XACML is the most commonly used standard for access control based on attributes. This standard describes the components, purpose, interaction and use of the access control system. The system is expected to apply to various types of data, such as multimedia


information, electronic documents, etc. It is not advisable to store this amount of data directly in the blockchain, as increasing the number and the size of the blocks increases the complexity of Ethereum, which mainly affects the cost of transactions. Therefore, data is stored in cloud storage in which the file identifying information is only available . The Ethereum platform is designed to create a blockchain-based decentralized service. It's a single virtual machine distributed. Smart contracts Unlike Bitcoin, Ethereum supports cycles that, on the one hand, have led to the introduction of fees for their implementation, called gas, and have significantly expanded their applications on the other. Changing the virtual machine status can be written in the full script language of Turing. For each file, the user creates a smart contract that stores owner information, access policy, hash sum of the s tored informat ion, c loud ident i fying information, and any changes to the file. Since the information stored in the blockchain is public, information must be encrypted before it is sent to storage and access controlled. The interaction scheme between the client, CA and AA is shown on Fig. a contract file is created to store data. It contains information on the location of the file in the cloud storage, access policy and information for additional owners. It is possible to interact with the file using the contract. The system supports four types of interactions: create, edit, read, and delete.

Figure 3. Access Control System

For altering the file's contact strategy, the CD does apprise of the contact environment and mechanisms of the ciphertext. It formerly modernizes the facts in the pact file and

substitutes the mechanisms of the ciphertext in the cloud . When removing a folder or a file, the pact case self-destroys and CD must eliminate it from the cloud. After erasing the file, the connection to it cannot be cast-off all over again in the scheme to eradicate the option of arguments. A worker wanting to read a file must bout the access strategy and have the essential keys to decrypt. After testing for strategy acquiescence, the worker obtains a connection to the file and can copy it, and then decode. If the user does not see access strategy, then the file it is to decode even if he will be capable to tie to it .

5. BLOCKCHAIN CLOUD AND RELATED SUSCEPTIBILITIES

Amongst all the safety matters that occur in the cloud atmosphere, blockchain will be very operative in addressing the tasks and challenges intricate in the application of certain data attribution. We extant the tests related to certain data attribution in the cloud and blockchain competences to report them.

A. Blockchain and Cloud Security

Cloud computing permits users to distantly stock their data into the cloud and delivers on-demand requests and facilities from a shared loch of configurable calculating properties. The sanctuary of the subcontracted data in the cloud is reliant on the safety of the cloud computing scheme and net. Though, cloud's key features, on-demand facilities, con t inuous ne twork con tac t , r e se rve assembling, and fast resistance are vulnerable to vulnerabilities. In adding, the cloud computing's central tools for virtualization, cryptography, and net amenities have susceptibilities, that are consequences of uncertain application. At a similar time, security checks, such as key organizations, in the cloud computing environment have numerous trials. For example, to the appliance, an operative key management scheme in cloud computing substructure needs administration and storing of many types of keys. The trouble in conveying standard key organization twigs from the point that simulated technologies typically have varied and heterogeneous hardware/software, and the cloud-dependent computing and storing are purely distributed.


The problem relies on the cloud substructure is that if some unauthorized entity tries to interfere and change the data, it cannot be detected. It happens because of the PKI based nature of the cloud. Therefore, a very strong attribution structure is required so that this problem can be eliminated and the responsible entity can be detected. Data authority is a thing that delivers information about all the changes and variations accomplished through data exchange between different units. Scholars have projected safety keys, such as PKI signatures, to guarantee the provenance. Whereas the application of PKI signatures normally rests on a central authority, that is not operative in the cloud substructure . Blockchain claims that it does not require a central system or central authority because its execution is different from the rest of the technologies. There are some ledgers which are distributed, that ledges hold and record all of the transactions and actions that have been done on data. After maintaining the record it shares that with all of the other users who are the participating units. Blockchain provides complete and safe transmission of information via a system of some cryptographically secure keys in a distributed environment. Hence, blockchain and keyless signatures can be the replacement of PKI signatures. The transactions in the public ledger are verified by a consensus of the majority of participating entities. The record of any transaction cannot be changed in blockchain technology because it is confirmable. Signatures that are keyless are those signatures that are unlike traditional digital signatures. These keyless signatures state an issue of “PKI key compromise”. PKI is “Public key Infrastructure” which depends on the distorted phenomenon of key cryptography. It disassociates the reliability and integrity protection and process from identifying the signer. These are those processes that are accountable for keeping the privacy and secrecy of the keys. The asymmetric cryptography and keyless cryptography are the choices from the techniques that are helpful in the identification of signer and protection of integrity. Hashing, publications, and aggregation are the methods of the keyless signature phenomena. One of the examples of keyless cryptography is “One-way Collision-free hash functions” The understanding of keyless signatures needs a Keyless Signature Infrastructure (KSI) that contains a pyramid of the co-operative aggregation servers which produce the universal

hash trees. The authentication in KSI centers on the safety of hash functions then accessibility of a community record (blockchain). The ledger is openly accessible and rules to bringing up-to-date, spreader consent and way of the process are well cleared.

I. DIFFERENCETable 1: Different Methodologies and their purpose

According to al l direct ly above deliberated techniques and methodologies of securing cloud storage with blockchain technology, the access control system has been the best approach so far. The Access control system uses the cipher-text policy to encrypt the data. The previously attribute-based encryption system used attributes to present encrypted data. While in cipher-text attributes they are used to define user's credentials. Text is plain before encryption and cipher-text encryption is the resultant data. The access control system also maintains a log of all events as shown in figure6 [5-9].


7. CONCLUSION

This paper presents a comprehensive approach to secure cloud storage using different blockchain methodologies. All the methods result in providing end-users to securely share their data. The best method among all of the above solutions is an Access Control System that provides a model to safe data by making it unchallengeable. The main idea is to adjust the access strategy for the encoded and encrypted data without repeating them to a huge amount of members that makes its presentation improved than ChainFS system and keyless signatures. Even though they have their individual compensations.

REFERENCES

[1] J. C. K. L. C. A. K. K. K. L. N. Qiwu Zou Yuzhe Tang, "ChainFS: Blockchain-Secured Cloud Storage," in 2018 IEEE 11th International Conference on Cloud Computing, New York, 2018.

[2] N. M. S. P. M. E. K. a. C. Y. Deepak Puthal, "The Blockchain as a Decentralized Framework," IEEE Consumer Electronics Magazine, p. 4, March 2018.

[3] S. Z. Ilya Sukhodolskiy, "A Blockchain-Based Access Control System for," 978-1-5386-4340-2/18/$31.00©2018 IEEE, p. 4, 2018.

[4] C. A. K. K. A. K. L. N. D. K. T. S. S. Xueping Liang, "Security Implications of Blockchain Cloud with Analysis of Block Withholding Attack," in 2017 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, 2017.

[5] T. Boxcryp to r, 2017 . [Onl ine ] . Available: https://www.boxcryptor.com/en/.

[6] H. M., "Attribute-Based Encryption Optimized for Cloud," SOFSEM, 2015.

[7] L. B. N. T. Courtois, "subversive miner strategies and block withholding attack in bitcoin digital currency," arXiv preprint.

[8] "Amazon AWS," [Online]. Available: https://aws.amazon.com/..

[9] G . D e v e l o p e r s , " G o o g l e c l o u d computing, hosting services & apis.," 2015.


Figure 4. Access Control System Mechanism Flow chart

IoT Based Architecture for Basketball SupervisionMubashir Ali, Sammya Hafeez, Mahnoor Khalid Paracha, Tehniyat Liaqat

1*,2,4Department of Software Engineering, Lahore Garrison University, Lahore, Pakistan3Department of Computer Science, Bahauddin Zakariya University, Multan, Pakistan1 [email protected], [email protected]

1. Introduction

Basketball is a sport with 360 to 450 basketballers in the current National Basketball Assoc ia t ion (NBA) and 144 to 180 basketballers in the Women National Basketball Associat ion (WNBA). The Basketbal l Tournament (TBT) is mainly a single-elimination tournament played each summer in America and Canada, currently consists of 64 teams and usually broadcast by Entertainment and Supports Programming Network (ESPN). These tournaments are basically responsible for increasing the progress and taking care of the condition of the game around the world. Jonathan Mugar was the founder of TBT. NBA has also introduced an innovation that has been made progress in the game. They introduce a technology which is the Replay System in which

alerts if a shot was discharged before the final buzzer or not. This system enables the couches to again view the footage of the game in good quality. This technology is implemented in 2002 in Western Conference finals. This system has made it very easy for the couches to make their last decision and it helps couches and players to see what they need to improve and where they made mistakes [1]. NBA is seeking ways to make Basketball a healthy sport too [2]. NBA players are known as the most paid players in the world by average annual salary per player. They also made an effort in the player’s health issues. It has made it compulsory to add full-time mental health staff for the basketballers. The given system of IoT basketball will benefit the instructor so they should be able to take the authentic decision in terms of the medication of an injured player by informing the health staff on

Abstract:

Basketball is one of the most played games in the world with a huge amount of fan following and has a great number of basketballers. Sometimes players get severe lower body wounds such as ankle sprains, shortage of breath, head, teeth, hand, and fingers. Female players have a higher risk of knee injuries than male players. These are health issues that players face while playing basketball. Sports organizations spend millions to train fresh basketball players or for the development of the previous basketball players. The internet of things (IoT) made everyday things readable, controllable and recognizable through the internet and the wireless sensor networks. It is simply the network of interconnected devices that are embedded with sensors, software, and connectivity modules. Nowadays, with this growing technology it is possible to protect the life of players in the game as well as in training sessions, if we detect the problems early in players and appropriate actions will be taken to reduce adverse health effects which can be very dangerous. In this paper, we will propose an architecture for basketball based on the internet of things (IoT). The main goal behind this approach is to introduce a healthcare system based upon sensors, actuators, devices and telecommunication technologies to communicating real-time stats.

Keywords: Internet of things, Cloud computing, Edge Computing, Basketball Architecture, Healthcare, Wireless Body Area Networks, Sensors


Ali et al LGURJCSIT 2019



30


time [3]. The Internet of things is referred to as actuators, sensors and any kind of embedded systems and these systems have the capability to generate any kind of data globally in an automated manner [4]. Many applications can be developed through the possibilities offered by the internet of things, in many diverse scenes, such as healthcare and wellness, homes and industrial automation[5], automotive, smart grids, etc [6]. The new generation internet enables global communication with the world using advanced communication technologies like 4G, 5G. Wireless networks providing a base to the internet of things for effective commun ica t i on . Among a l l enab l i ng technologies of the internet of things, there are two such innovative systems that can be useful for development such as Radio Frequency Identification abbreviated as (RFID) and Wireless Sensor Network abbreviated as (WSN) presents two of the very most up and coming solutions. IoT providing mechanism to gather data from RFID or mobile devices, and sensors. RFID is a technology mainly used to record data, control the targets and it can also identify objects. RFID is an electronic chip-based technology or a system for identification. In this technology, any device can be controlled by radio waves [7]. Then it transfers the data to the first part and then to the second part. Moreover, telecommunication is generally sending or receiving the information by the wireless or non-wireless medium. The nodes of IOT must have the ability to communicate with another object such as machine and humans and this should be the machine to machine habitat. Internet of things can be enforced in many of the applications such as care of health [8], smart cities [9], agriculture [10], smart grid [11], smart building [12], home automation [5], saving energy, and many more. It enhances the lifestyle of individuals and society. A lot of development work has been done on the internet of things in the past. That work is reliant on the tools or methods, classifications, components and these are needed in a general place. ETSI’s M2M is one of the most popular IoT standards. It has much recognition in terms of automated systems [11]. There are two parts to this standard. The first part consists of away, and its work is to collect important data from the devices of the system [11]. The other part is the connection part and executive, which consists of these factors such as storing data, accessing information

securely, analyzing and routing [13].

Figure. 1. IoT in Sports

Best technologies for examining human stats are Wireless Body Area Networks (WBANs) [14]. It can determine important physical functions such as heart rate, the motion of body and blood pressure [15]. Thus, it is possible to form a health system that keeps an eye on the health of a basketballer using automation which is based on the internet of things, and it is able to reduce the different health issues which can cause a negative effect on basketballers. In this new generation, the sensing devices enable us to connect to the networks by using interconnection technology [16]. Furthermore, IPv6 is placed by 6LoPWAN ( L o w p o w e r w i d e a r e a n e t w o r k ) o n comparatively small devices so the data is transferred with strong packet transportation and low power utilization. We will introduce IoT Basketball, and its purpose is to attentively observe basketballers when they are playing, and if unusually any injury or accident occurs the problems can be recognized and solved on time [17]. It basically works by putting heat gadgets on the basketballers when they are playing in the tournaments or in trails to instantiate essential frameworks mainly temperature of the body, and rates which include sweat, heart, body, and respiration [18]. In section 2 we will talk about the technology used in this application and it includes the factors which are required for this architecture. In the 3rd section, we will discuss a case study with two cases that highlight the importance of this architecture and finally propose IOT based architecture with four layers perspective. 4th section is all about the


illustration of some difficulties that could be encounter while implementing this approach in practice. Lastly, in the 5th section, we conclude our work and provide guidelines for further improvements.

2. LITERATURE REVIEW

Nowadays, technology has become the most important part of people’s life. It has made life so much easier, safe, reliable and convenient by providing an enormous amount of services. Wearable monitoring devices play a major role and are very helpful in monitoring and collecting physiological data such as blood rate, daily activity and oxygen level which improves the quality of life. In the future, the health care system will be very improved with the use of these devices because these can detect disease early on and prevent it from occurring by providing proactive wellness management. The most important technology which is developed recently and also implemented in the health care system is Wireless Body Area Networks [18]. Most of the diseases occur because there is no time-based checkup for the players in order to get quick treatment by reporting it in the first place. Wireless Body Area Networks (WBANs) help the players to get the treatment accordingly and they don’t have to stay in hospitals for a long time. It consists of actuators and small bio-medical sensors that are found on the patient’s body. Sensors measure the body parameters and actuators work on the data received by sensors. This technology allows monitoring the temperature of the body, heartbeat and blood pressure of the player. With the help of Internet Communication Center, this data is sent to electronic devices such as phones, laptops, and tablets. Data is gathered by the sensors [19]. WBANs provide the possible solutions and are pro-active, very fast and comparatively affordable as compared to the other health monitoring systems. WBANs are the core technology that is used to report any abnormality of the human body and to diagnose any kind of disease [20]. After analyzing the data, it also tells the required solution or treatment of the problem. Moreover, WBANs are non-invasive technology and economical solutions for any healthcare application. Other wireless devices such as WSNs, WPAN, MiWi, ZigBee, WLAN along with the internetwork with the WBANs.

WPAN and WLAN abbreviated as Wireless Personal Area Networks and Wireless Local Area Networks. WBANs are an emerging technology with the aim to provide a comfortable life by providing suitable care systems for its patients. Many health care systems are developed on the basis of WBANs and research efforts are also in progress to enhance this technology. There are other wireless networks too which are exploring in the health sector such as Vimax and wireless body sensor networks. The main objective of this paper is to develop a system for health care which's purpose is to report the player’s disease early so the players can not suffer from it. There is a lack of attention towards the health care system for the players in the sport that is why we are proposing this Basketball Architecture based on IoT. However, many research articles proposed on this by other models. [21] Presents a monitoring system that can be used daily and also in the field of sports. This system consists of body nets and totally dependent on body contact. It can measure many body diseases such as the amount of oxygen, heartbeat, sweat rate and blood pressure which can be found by photoplethysmography signal. When data is received, it gives solutions to the people who are in the field of sport as well as to the people who d o d a i l y p h y s i c a l e x e r c i s e . Another article on the health care system is proposed by [22] in which a personal computer application software is implemented and is based on WBANs. These computers work on the embedded sensors which are used to monitor body movements and ECG sensors to see heart activities. These two systems worked so well but there is one limitation that is they can only work in a specific environment. And develop a system that is convenient for the players while playing and training. Upcoming researches have shown interest in developing a system for the players which is light-weight and cannot irritate the player while playing such as bracelets, headbands, shorts, and T-shirts. With the emerging technology people prefer WBANs for health care monitoring as it is more sufficient and reliable. [23] introduced a device that is a shoe-based wearable device and its purpose is to measure different postures of the human body such as standing, cycling, jumping, walking and running, taking steps in ascending or descending manner and helps the people suffering from obesity. Furthermore, a system, [24] was proposed which works with pH monitoring


systems having pH sensors and it measures the amount of sweat of the patient in order to see the problem. [25] Introduced wearable shirts. These shirts measure the player’s ECG signals and its activities in order to detect any unusual activity in the player’s body. As many research papers are focusing on developing a system on health care in sports and many of them are already introduced in different terms by using different technologies. [26] Introduced a system using IOT technology for patients in which they can stay at home such as an easy environment for them. It contains three parts mainly sensors, Microsoft XBOX and ICU monitoring system. The writers of [27] introduced an architecture based on IOT to build a system based on health care using ETSI's M2M standards. Bluetooth has been also introduced as WPAN protocol and is very famous for its large connection availability for current medical systems. Yet there is a shortage of research on introducing or making a system for the sports in order to give facilities for monitoring which is based on the architecture of IoT. IoT Basketball provides a system that prevents the players from any kind of incident, sports-related risks, and major injuries. Basket baller’s activity can be seen with this system by measuring different parameters and also measuring the parameters of the area such as intensity and temperature of that place. This work can be done by using different technologies and sensing devices which can be placed on the player’s kit or by installing it under the player’s skin or in his tissues. This approach enables the communication with Constrained Application Protocol and IP address. These technologies can be accessed by the web services. RFID is implemented in the system for the basket baller’s identification and the Routing Protocol for Low Power (RPL) is implemented in IOT basketball.

3. PROPOSED ARCHITECTURE

3.1 Case Study

As we already know basketball is one of the most played games in the world with 2 million fans and 450 plus professional basketball players from all around the world. It is played in the rectangular court and you scored the points depending on which side you throw the ball into the hoop. The ball can be moved in the court by plopping and passing the ball to the other players. The end result of the game is

declared by the referees and the team with the most points will win the game. The teams will consist of 12 players and only 5 are allowed to play in the court. There are different positions by the players such as center, defensive forward, offensive forward, defensive guard and point guard. Each player then will take his positions but also allowed to move in the court as per requirements [28]. A case study is being illustrated in this section that how IoT Basketball will help the basketballers in terms of their health issues which occurs during the game or training such as accidents or sudden illness. We will briefly talk about the problems and see their solutions. Figure 2 shows the basketball court with dimensions. Proposed IoT based architecture for basketball which describes the technologies that will be enforced to provide the players with safe and effective outcomes which leads them to the higher levels of performance and quick recovery which ensures the success of players, while making sure the players are healthy and fit as they were before. Taking in view the case study of a first player. First Player’s name is Charlie Puth. Charlie has been in this game for more than nine years and is treated as the professional in the game. In his game history, there is no clue of any major illnesses in his medical background. While he was playing in the field one day, he became unconscious. There can always a risk w i t h a l l t h e p l a y e r s t h a t o t h e r t h a n unconsciousness they can face any health issue on the field while playing the game. It includes heart attack, ankle sprains, fatigue, eye to eye happenings and many other small injuries. Our system will help the coaches and supervisors to monitor the health issues of players before the game and even in the field without any delay because our technology operates immediately.

Figure. 2. Basketball Court


Now talking about the case study of another player. The second Player’s name is Will Smith. Smith has been in this field for almost five years and he has never been sick on the field. One day suddenly, he suffered from a minor heart attack. This happened because there was no system to measure the health of the players in the field or before the game. In the next section, we will illustrate how this technology will help in preventing the situations and how it responds if the player faces any major health issue. Firstly, [29] we used RFID technology to get personal information about the player, which includes his good name, registration number, designation, his medical background and he plays with which team. The environment of the court along with the player's condition will be determined. These include mainly the temperature of the court, the intensity of light, body temperature, heartbeat, level of oxygen, sweat rate, blood pressure, and respiration rate. We will use the default gateway to transform the information or the signals to the Adafruit cloud. Adafruit cloud has many built-in functions and libraries which are available for health care monitoring. We are using the MiWi protocol for this instance. After Adafruit cloud analyzed the information, a notification is sent to the coach or the team supervisor as a warning by an electronic device that tells them that Charlie in first case and smith in the second case is not okay, and they need medical assistance immediately and take them out of the court. They are provided with medical help. If after the recovery something happened to any of them the technology will be operated in the same way. Our system will make sure that the players will be alright and they are doing well and we do this by monitoring them. This system will get to know the condition of the player even before himself and how to fix or prevent it from happening. It means that there will be a minimum possibility of any injury or major health issue while he was focusing on his game.

3.2 IOT Basketball Architecture

IoT a l lows di fferent devices to communicate with each other via the internet or wireless networks. To support the case studies we mentioned above we also proposed an IoT architecture which is shown in figure 3. We used different sensors that are lightweight, got their own IPv6 capabilities (6oLWPAN) and have low power consumption. It measures the physical


parameters such as sweat rate, heartbeat, motion, and temperature of the body, breathing rate, and blood pressure. This information is sent to the cloud through wireless sensor networks with the gateway. To measure the environment of the court we use other sensors too which includes the temperature and humidity in the court. This layer in the system is known as the perception layer. To communicate with the default gateway the Constraint Application Protocol (CoAP) is used in the sensors [30].

Figure. 3. IOT Basketball Architecture

Th i s p ro toco l i s used to avo id conjunction between the technologies and it makes a gateway to integrate with the web by using HTTP. Furthermore, CoAP has many advantages such as multicast support, low header overhead, asynchronous message exchange and it is based on the protocol which is Datagram Protocol (UDP) which is used to transport with an application layer. The second block in the figure referred to as a Network layer that collects all the data from the perception layer and transfers over the cloud. MiWi protocol is the default communicator which is used in the IoT basketball architecture. Data collected from the perception layer is transferred over the cloud using the MiWi base station. The ETSI’s M2M is divided into 2 parts. ETSI’s M2M NSC is used for the network application. Network application works for the registration, security, routing, and NAT. MLA interface is a bidirectional data flow is provided

by MLA Interface and it works as a passage between the network application and NSC. To perform IOT Basketball in the world there are many functions and characteristics that a system must operate. The next section provides the details used in every part. The basketball IoT architecture consists of four layers that are under:

1. Perception Layer2. Network Layer 3. Edge Computing4. Application Layer

A. Perception Layer in IoT Basketball Architecture

This layer consists of sensors having two networks. We use the wireless body area networks (WBANs) to measure the health condition of the basketballer. These kinds of sensors must be lightweight because players need to wear them such as shirts, Bracelets, and shorts. To communicate with the gateway of the system, every sensor must have its own IPv6 (6LoWPAN) and must be working with Routing Protocol for Low-Power (RPL). The CoAP Protocol works with the sensors to send and receive the information. Moreover, the devices must fulfill the security capabilities and QoS requirements. The devices around the court which measure the temperature and humidity do not have the same capabilities as the body sensors, but they need to cooperate with the body sensors in order to transmit data over the cloud computing platform known as adafruit cloud.

B. Network Layer in IoT Basketball Architecture

In this system, the MiWi protocol is used for the base station. The base station must be connected with a network domain with a wireless connection. The base station also needs to work with the 6LoWPAN, which means supporting the IEEE 802.11 standard of wireless communication. Moreover, the default gateway must implement the CoAP Protocol to send and receive the information from the sensors. It enables QoS Requirements and security efficiency [31].

C. Edge Comput ing Layer in IoT Basketball Architecture

Edge computing refers to end node based infrastructure that processes instant data in real-time fashion for quick response [32]. In layer three, the sensed data is processed and notified to supervisors before sending it to overcloud. The complex operations are performed overcloud. It also performs data-handling activities or other network operations. It performs the data analysis instantly and presents the results to the users. It delivers the data to the endpoint devices such as mobile phones, laptops, and tablets. It performs computation on the edge devices. Moreover, these devices move large amounts of processed data it improves the latency, quality of service, reliable and quick response and cost of transmission.

D. Application Layer in IoT Basketball Architecture

The Cloud computing service known as the Adafruit cloud is used over the application layer [33]. It examines traffic in the data that receives from the technologies when the conjunction occurs in order to facilitate the coach or the superintendent of the game. To analyze the injuries and health conditions of the basketballers we can possibly use predictive analytics in data mining. Adafruit cloud services ensure the quality of service and security efficiencies. The information of the players should be stored in a secure place. Furthermore, the Adafruit cloud is more suitable as compared to the other software and used to provide the correct information to the superintendent of the game. It should be able to give complete information about the basketballer such as health conditions, preceding ailments, and weaknesses.

4. DISCUSSION

We c a n s e v e r a l h u r d l e s w h i l e implementing this system for basketball which is as follows:

• Some players don’t like the idea of their private information being gathered by the sensing devices because their privacy is more important and if it got leaked to the other people it can cause damage and have an adverse effect on their life.


• Any communication system will need to be upgraded as the passage of time, technology will be evolving in performance levels, the quality of gadgets, speed of processing and many more. For this, we should be updated about the new happenings. • In order to implement the required technologies in the court and the gadgets, this whole process can be very expensive as it changes the whole structure of the court.• There can be technological errors such as the coach does not receive instant notifications edge computing layer and the connection between the sensors and the default gateway.• To manage all the gadgets, automation and structure of the system to the high level.• The ability of computer systems and software to work perfectly is crucial to effectiveness. So they have to work with flexibility in order to function properly and to provide accurate results.• IoT basketball consists of different sensors and layers in order to work effectively between the basket baller’s wearable devices in the body and to the superintendent electronic medium. It makes an effective interaction between both.• Many architectures of IoT have been presented before to find a relevant standard to form an IoT Basketball system and it is essential for us to find one above of the success.

5. CONCLUSION

I o T r e f e r s t o i n t e r c o n n e c t e d communication between actuators, sensing devices and embedded systems. It strengthens the style of living in a society via effective automation. It has much utilization which consists of smart cities, health care, smart buildings, home automation, intelligent traffic system and many more. In this research, we have proposed a basketball architecture, which aims to monitor basketballers while they are playing or in the trainings so if any injury or accidents happened in the game it can be overcome and given the solutions. This can be made possible by putting the wearable sensing devices on basketballers during the game and transmits all the information to the application layer for storing and processing of the data, and then finally sends information to the superintendent with the help of electronic devices such as phones, laptops and tablets so they can see if

anything happened to the player in the court or not. The architecture of this system uses ETSI’s M2M standard. In the future, we will enhance this architecture for other sports such as football, volleyball to prevent health issues. TinyOS simulation will be used for the implementation of this system and to address any kind of circumstance that can occur.

CONFLICT OF INTEREST

O n b e h a l f o f a l l a u t h o r s , t h e corresponding author states that there is no conflict of interest.

Acknowledgment

We would like to thank journal editor, area editor and anonymous reviewers for their valuable comments and suggestions to help and improve our research paper.

REFERENCES

[1] B. Chettri, S. Mishra, B. L. Sturm, and E. Benetos, “Analysing the Predictions of a CNN-Based Replay Spoofing Detection System,” in 2018 IEEE Spoken Language Technology Workshop, SLT 2018 - Proceedings, 2019.

[2] C. E. Gross, J. Huh, and J. A. Nunley, “Basketball,” in Foot and Ankle Sports Orthopaedics, 2017.

[3] M. A. Ikram, M. D. Alshehri, and F. K. Hussain, “Architecture of an IoT-based system for football supervision (IoT Football),” in IEEE World Forum on Internet of Things, WF-IoT 2015 - Proceedings, 2015.

[4] T. Jensen and M. Durham, “Internet of things,” Advancing Microelectronics, 2017.

[5] M. Alaa, A. A. Zaidan, B. B. Zaidan, M. Talal, and M. L. M. Kiah, “A review of smart home applications based on Internet of Things,” Journal of Network and Computer Applications. 2017.

[6] J. Lin, W. Yu, N. Zhang, X. Yang, H. Zhang, and W. Zhao, “A Survey on Internet of Things: Architecture, Enabling Technologies, Security and Privacy, and Applications,” IEEE Internet Things J., 2017.


[7] S. Amendola, R. Lodato, S. Manzari, C. Occhiuzzi, and G. Marrocco, “RFID technology for IoT-based personal healthcare in smart spaces,” IEEE Internet Things J., 2014.

[8] G. Kaur, P. Tomar, and P. Singh, Internet of Things and Big Data Analytics Toward Next-Generation Intelligence. 2018.

[9] Z. Kamal, A. Mohammed, E. Sayed, and A. Ahmed, “Internet of Things Applications , Challenges and Related Future Technologies,” World Sci. News, 2017.

[10] A. K. Saha et al., “IOT-based drone for improvement of crop quality in agricultural field,” 2018 IEEE 8th Annu. Comput. Commun. Work. Conf. CCWC 2018, vol. 2018-Janua, pp. 612–615, 2018.

[11] L.-Y. Chen and C.-Y. Huang, “An open internet of thing architecture integrating OnEM2M and OGC sensor th ings API standards,” in Proceedings - 39th Asian Conference on Remote Sensing: Remote Sensing Enabling Prosperity, ACRS 2018, 2018.

[12] I. Rosadi and S. P. Sakti, “Low-cost wireless sensor network for small area in a building,” in Proceedings - 2017 International Seminar on Sensor, Ins t rumenta t ion , Measurement and Metrology: Innovation for the Advancement and Competitiveness of the Nation, ISSIMM 2017, 2017.

[13] G. Katsikogiannis, D. Kallergis, Z. Garofalaki, S. Mitropoulos, and C. Douligeris, “A policy-aware Service Oriented Architecture f o r s e c u r e m a c h i n e - t o - m a c h i n e communications,” Ad Hoc Networks, 2018.

[14] K. C. Karthika, “Wireless mesh network: A survey,” in Proceedings of the 2016 IEEE Internat ional Conference on Wireless Communications, Signal Processing and Networking, WiSPNET 2016, 2016.

[15] . K . L . K . , “ B O D Y A R E A NETWORKS,” Int. J. Res. Eng. Technol., 2016.

[16] “Evaluation of Wireless Body Area Networks,” Int. J. Innov. Technol. Explor. Eng., 2019.

[17] S. Al-Janabi, I. Al-Shourbaji, M. Shojafar, and S. Shamshirband, “Survey of main challenges (security and privacy) in wireless body area networks for healthcare applications,” Egyptian Informatics Journal. 2017.

[18] M. Ghamari, B. Janko, R. S. Sherratt, W. Harwin, R. Piechockic, and C. Soltanpur, “A survey on wireless body area networks for e h e a l t h c a r e s y s t e m s i n r e s i d e n t i a l environments,” Sensors (Switzerland), 2016.

[19] S. Majumder, T. Mondal, and M. J. Deen, “Wearable sensors for remote hea l th monitoring,” Sensors (Switzerland). 2017.

[20] I. Lourenço and J. Costa, “In-shoe sensor system with an embedded user interface and wearable leg unit,” in 6th IEEE Portuguese Meeting on Bioengineering, ENBENG 2019 - Proceedings, 2019.

[21] Y. Fu and J. Liu, “Monitoring system for sports activities using body area networks,” in BODYNETS 2013 - 8th Internat ional Conference on Body Area Networks, 2013.

[22] C. Otto, A. Milenković, C. Sanders, and E. Jovanov, “System architecture of a wireless body area sensor network for ubiquitous health monitoring,” J. Mob. Multimed., 2006.

[23] E. S. Sazonov, G. Fulk, J. Hill, Y. Schutz, and R. Browning, “Monitoring of posture allocations and activities by a shoe-based wearable sensor,” IEEE Trans. Biomed. Eng., 2011.

[24] D. Morris et al., “Wearable sensors for monitoring sports performance and training,” in Proc. 5th Int. Workshop on Wearable and Implantable Body Sensor Networks, BSN2008, in conjunction with the 5th Int. Summer School and Symp. on Medical Devices and Biosensors, ISSS-MDBS 2008, 2008.

[25] Y. D. Lee and W. Y. Chung, “Wireless sensor network based wearable smart shirt for ubiquitous health and activity monitoring,” Sensors Actuators, B Chem., 2009.

[26] I. Chiuchisan, H. N. Costin, and O. Geman, “Adopting the internet of things technologies in health care systems,” in EPE 2014 - Proceedings of the 2014 International


Conference and Exposition on Electrical and Power Engineering, 2014.

[27] P. Diogo, L. P. Reis, and N. V. Lopes, “Internet of Things: A system’s architecture proposal,” in Iberian Conference on Information Systems and Technologies, CISTI, 2014.

[28] B. Alamar, “Basketball on Paper: Rules and Tools for Performance Analysis,” J. Sport Manag., 2016.

[29] Kubo, “The research of IoT based on RFID technology,” in Proceedings - 7th International Conference on Intelligent Computation Technology and Automation, ICICTA 2014, 2015.

[30] I. J. Shin, D. S. Eom, and B. K. Song, “The CoAP-based M2M ga teway fo r

distribution automation system using DNP3.0 in smart grid environment,” in 2015 IEEE International Conference on Smart Grid Communications, SmartGridComm 2015, 2016.

[31] M. A. Razzaque, M. T. Hira, and M. Dira, “QoS in body area networks: A survey,” ACM Transactions on Sensor Networks. 2017.

[32] W. Yu et al., “A Survey on the Edge Computing for the Internet of Things,” IEEE Access. 2017.

[33] R. Gayathri and S. K. Vasudevan, “Internet of things based smart health monitoring of industrial standard motors,” Indones. J. Electr. Eng. Informatics, 2018.


Parallel Data Transmission using New Line Encoding Method 1 2Waqas Ahmad , Awais Salman Qazi

1Departement of Computer Science & Information Technology, Lahore Garrison University, Lahore, Pakistan2Department of Computer Science & Information Technology, Lahore Garrison University, Lahore, Pakistan

1 2Email: [email protected] , [email protected]

1. Introduction

I n c o m p u t i n g e n v i r o n m e n t s l i k e telecommunication and networks field, a line coding which is also known as digital baseband, the modulation is selected to be used in digital communication and transmission environment for a purpose of baseband transmission. There are various line coding algorithms that are widely used for digital data transmission in which binary data in the form of one (1) and zero (0), is represented in various digital signaling formats. The same is the case in Pulse Code Modulation signaling [2, 6]. In every encoding method, binary data is sent using a number of rectangular-shaped pulses. The decision of what size of pulses should be used to represent the binary bits one (1) and zero (0) is normally made on the ground realities using the following considerations i.e. whether DC level is present or absent, efficiency of bandwidth, level of transparency, is it easy to recover clock signal and availability of error detection characteristics

[2, 6]. There are many line codes available right now, the most popular line codes which are in use worldwide are named as unipolar encoding, polar encoding, bipolar encoding, Manchester encoding and differential Manchester encoding. Brief and precise introduction of popular codes are written as below.

1.1 Unipolar Encoding:

In the unipolar line coding, binary bit one (1) is shown by an upward positive voltage, and a binary zero (0) is represented by a horizontal line showing zero voltage level. This line coding is the simplest technique. Another name of this technique is on-off keying, termed as OOK, and also known as NRZ (nonreturn to zero) scheme. This coding method is preferred when one symbol is sent much more often as compared to the other. However, this coding technique has some deficiencies as well which are, it is not self-clocking and normalized power is double as compared to the polar NRZ. Therefore, in

Abstract:

In our research paper, we have introduced a unique and new line code method/technique that will have the capability to send two different codes in parallel or concurrently using a single line code, and this will be achieved without any major loss in the shape of the signal. In our proposed new line code, the structure of our code is based on two already well-known line codes that have been used many times in the data communication field, both of them have been combined together. We can say that our new line coding method is a hybrid encoding scheme. This line code combines the properties of already existing line code techniques. At the receiver end, we have implemented a very easy and basic separation technique that shall help us to separate the real user codes from the added line code, and this will happen without any disturbance or major loss/distortion. There will be no effect on the signal of input data.

Keywords: Line Code, Distortion, Parallel


Ahmad et al LGURJCSIT 2019



39


present days, unipolar encoding is not used in the data communication networks field [1].

1.2 Polar Encoding:

The polar encoding uses two levels; these levels are either known as polarity or amplitude. It is categorized into three types i.e. NRZ (not return to zero), RZ (return to zero) and biphase. Binary number 1 is represented by the positive voltage level while binary 0 is represented by the negative voltage level [1].

1.3 Bipolar Encoding:

Bipolar encoding is a type of RZ line coding, in which we use two non zero values, so the three values are plus (+), minus (-) and zero (0). Bipolar signaling is often called a duobinary signal. Binary bit 1 is represented by positive voltage and then negative voltage while binary bit 0 is represented by a constant zero value. One of the various benefits of bipolar encoding scheme over unipolar is error detection. The term pseudoternary also belongs to bipolar signaling, it refers to the use of three encoded signal levels to represent two-level binary data. Another name is alternate mark inversion [1, 7].

1.4 Manchester Encoding:

Manchester encoding is a subtype of po la r b iphase d ig i t a l cod ing . In t he telecommunication field, Manchester encoding is a type of digital encoding in which binary one (1) is represented by a positive half voltage level then represented by a negative voltage level. Similarly, binary bit (0) is represented by a negative half voltage followed by a positive half voltage level. This line coding technique is different than others in which a bit is represented by +5 volts high state and low state is represented by 0 volts [5, 8].

1.5 Differential Manchester Encoding:

This encoding scheme is also a subtype of polar biphase encoding. It is the technique in which data and clock signals have been combined together to form a single level 2 data stream. To indicate the logical values, presence or absence of transitions are used in this method. Synchronization is much easier in this method as compared to other encoding methods because the only polarity at the receiver side matters,

whether polarity received is the same or different from the previous value. It doesn’t matter what logical value is reaching at the receiver side. This line code method is known by many names like Biphase Mark Code, Frequency Modulation, Aiken Biphase, and Conditioned Diphase. In this encoding method, it is not necessary to know the polarity of the transmitted signal. This encoding method has various advantages over already present line code methods like it has robust clock recovery, detection of transmissions is less error-prone and it has zero DC bias. In differential Manchester encoding, there are two versions of the code, in which one version makes the transition for binary bit 0 and no transition for binary bit 1, similarly, another version of code makes the transition for binary bit 1 and no transition for binary bit 0 [6, 9]. Differential Manchester NRZ line code has the advantage of always having a 0 DC value, it doesn’t matter what kind of data sequence will be, most importantly it has twice the bandwidth as compared to the unipolar NRZ and polar NRZ code because the size of pulses is half of the width [5]. The waveforms of different line codes are shown in figure 1 below as an example when the stream of data bits is (01101001). All Waveform shapes are different from each other because encoding methods are different in each case. Following encoding, techniques are applied on the given binary data, i.e. Unipolar not return to zero encoding method, polar not return to zero encoding method, unipolar return to zero encoding method, bipolar return to zero encoding method and Manchester encoding method [1, 18]. We can easily see that stream of bits’ representation in the form of waveform is different.


Fig 1. Signal Representation using different encoding techniques

2. STRUCTURE OF PROPOSED ENCODING SCHEME

In our manuscript, we have introduced a unique and different line code technique that can be used to transmit two codes in parallel at the same time using the same bit interval of the one-bit period. And particularly in this section, we have tried to show the structure of the proposed new l ine cod ing t echn ique based on distinguishing features. However, conventional line code techniques only have the tendency to send one code per bit, which is more time taking and most error-prone [3].

2.1. Shapes of Signal using New Line Code

In this manuscript, we have shown how parallel transmission can be done using a new method. We take two bits from two separate codes i.e. the first bit is taken from the first line code and the second bit is taken from another line code. By this coding strategy, we will have four combinations of data bits that are to be transmitted. Let’s suppose we have 2- bits input whose combinations are (00), (01), (10) and (11). Now if we want to show the above bits by using our proposed line coding scheme, it will be represen ted in a d i ffe ren t way. Th i s representation will be done as follows.

Table No. 1. Data Signals using New Line Coding

By looking at the above table we can observe the shapes of the signals after they have used new line encoding. The signal is having unique characteristics that are described in bullets below.

• When the data bits are (01), the 0 bit is shown by the horizontal line and the 1 bit is shown by negative pulse downwards.

• When the data bits are (10), the 1 bit is shown by a positive pulse upwards and bit 0 is shown by the horizontal line.

• When the data bits are (00), the first zero

is shown by a negative amplitude pulse and the second zero is shown by a positive amplitude pulse.• When the data bits are (11), the first 1 bit is shown by positive pulse upwards and the second 1 bit is shown by negative pulse downwards.

2.2. Transmission Strategy

We have summarized the transmission strategy of our new line code technique. The way binary bits are transferred using new line method is explained in simple steps.• Two separate codes will be combined together bit by bit in such a way that the first bit belongs to the first code and the second bit belongs to the second code.

• Represent each and every combination of code by its relevant pulse shape and then transmit the signal to the receiver side.

Let us take a look at a scenario of transmitting two separate codes, which will give a clearer idea that how the data is sent using our new line encoding technique. Assume the binary codes are (01010) and (11000) as an example which we want to transmit using new line encoding method. Table 2 shows the two different binary numbers and its converted new line code data. Converted new line data will be transmitted.

Table No. 2. Data encoded using New Line Code

The signal that will be transmitted using new line coding will look like in the following shape, which is shown in figure 2. The upper wave shows a positive voltage level while the lower wave shows a negative voltage level. The following waveform can also be verified if we take a look at table 1 again.

Fig 2. Waveform after using new line coding


3. PERFORMANCE OF LINE CODE

In this section, we have compared our proposed new line code based on performance with all already existing line codes techniques and the parameters we discuss here are self-synchronization, the density of the power spectral and the decoding algorithm used for new line coding.

3.1. Self Synchronization

It is the Line coding technique that will make it possible for the receiver to synchronize the phase of the signal received. For instance, the synchronization of the signal is not done according to the needs, then the decoded signal will be problematic and not reach to the desired optimal differences in terms of the amplitude of different digits in line code. Due to this, there will be huge chances of error in the received signal [15]. Our new line code method is different, and possesses a feature that works on a technique of one-bit transition, therefore our new line code is better than all existing line codes. Our new line code method has features like self-synchronization which is helpful in detecting errors.

3.2. Relationship of PSD and Normalized Frequency

The power spectral density (PSD) function exhibits the strength of energy as a function of frequency [6, 16]. It is worthy to note that PSD of any line code depends on the shapes o f p u l s e s a n d r a t e s t h a t h a v e t o t a l correspondence with the digital values. As we already know that our proposed new line coding technique is a combination of two different codes which are Manchester encoding and polar RZ encoding. Therefore, the PSD of our proposed new line coding will be the average of the PSD of two other codes named Manchester Encoding and Polar RZ Encoding [17]. Below figure number 3 shows the PSD of our proposed new line coding technique.


Fig 3. PSD w.r.t Normalized Frequency

In figure 4 below, we have shown the comparison of the PSD of already existing line coding techniques with our proposed new line coding technique in the case of parallel data transmission. We wanted to be sure whether our proposed line encoding technique shows better results in terms of PSD, and the below figure clearly indicates that New Line Coding's PSD is best of all. It can be seen in the below figure that our proposed new line coding technique outclasses all line code techniques and possesses the properties that remaining line coding techniques don't have it. For example, new line code technique outclasses others in terms of smaller DC values and less spectral contents [11, 12].

Fig 4. PSD Comparison of Various Encoding Methods

4. Decoding Algorithm of Proposed New Line Code

The decoding algorithm used in the new line code technique is very simple and effective, which we have already discussed in previous sections. Once the data is encoded at the transmitter side, it has to be decoded at the receiver side to obtain the real data. Every encoding method has its own way of decoding the transmitted signal. We have used a different way to decode the transmitted signal. Simply, we have to use one-line code for one bit and another line code for upcoming bit, and then combine together. This algorithm is explained in steps as follows.

• In the new line scheme, the code of first bit always refers to the positive side of the new line code while the code of the second bit always refers to the negative side of new line code.• In the new line scheme, the top upper side if the user input data is binary bit 1 then the positive value will be in the first half of the bit

period and on remaining half will be negative.

• In the new line scheme, the lower downward side if the user input data is binary bit 1 then the negative value will be in the second half of the bit period and on remaining half will be positive.

5. Conclusion

The new line code technique shown in this research paper has shown wonderful performance in terms of clock recovery, phase spectral density, and bandwidth efficiency. All existing conventional line codes have also been checked with these parameters and it is clearly observed that our proposed new line coding technique is far much better than others. None of the existing line coding techniques have the tendency to send two codes concurrently, but the new line code technique can send two codes concurrently, which is a huge plus point. By doing this one could conserve the bandwidth. In a computing environment, it is very necessary that we use an efficient encoding technique. The new Line coding scheme has many advantages over previous encoding techniques. This proposed line coding technique scheme will contribute a lot to the telecommunication and networks field.

REFERENCES

[1] Khmaies Ouahada, Hendrik Christoffel Ferreira, “Simulation Study of the Performance of Ternary Line Codes under Viterbi Decoding”, IEEE Proceedings Communications, Vol. 151, No. 5, November 2004.

[2] Micheal P. Fi tz , July 18, 2008, “Fundamentals of Communication Systems”, 1st edition, McGraw Hill Education.

[3] Valeriu Munteanu, Daniela Thiceriu, “Information Analysis of the NRZ-M Line Codes and Their Generalization”, International Symposium on Signals, Circuits and Systems, Romania, IEEE, 2005.

[4] Zuojian Song and Yoshitaka Takasaki, “Line Coding for Clock Recovery with Minimal Filtering”, IEEE Pacific Rim Conference on Communications Computers and Signal Processing, Victoria BC Canada, 2007.

[5] Nikola Alic, “Performance Benefits of Line Coding in the Context of Direct and Coherent Detection”, Lasers and Electro Optic Society Annual Meeting-LEOS, November 2008.

[6] Simon Haykin, “Digital Communication Systems, John Wiley & Sons”, Inc, Feb 2013.

[7] Bernard Sklar, Pabitra Kumar Ray, “Digital Communications Fundamentals and Applications”, 2nd Edition, Pearson Education.

[8] M. Bhagyaveni, R. Kalidoss and K. Vishwaksenan, “Introduction to Analog and Digital Communication”, River Publishers Series in Communication, Vol. 46, Year 2016.

[9] Simon Haykins, “Communication Systems”, John Wiley & Sons, 2016.

[10] Dae Young Kim, “Condition for Stable Minimum-Bandwidth Line Codes”, IEEE Transaction on Communications, VOL. COM-33, NO. 2, Feb 1985.

[11] H. Chung, S. Jeon, “Clock and Data Recovery of an Extended Manchester Code for Pulse Amplitude Modulation”, 978-1-4673-4828-7/12, IEEE 2012.

[12] D. oner, “Criteria for Choosing Line Codes in Data Communication”, Istanbul University, Journal of Electrical and Electronics Engineering, Vol 3, No. 2, 2003.

[13] Swa t i Ve rma and Roh i t S ingh , “Multilevel NRZ Coding for Transmission of Digital Signals”, International Journal of Advanced Research in Science and Engineering, Vol. No.2, Issue No.9, September 2013.

[14] Nuha A and Abdelrasoul J, “Method of Unipolar Digital to Digital Encoding Data Transmission”, Journal of Engineering IOSR, Vol. 04, Issue 03, March 2014.

[15] Glass A Bastaki, “H-Ternary Line Code for Data Transmission”, Internat ional Conference on Communications, Computer and Power ( ICCCP 2001) , Sul tan Qaboos University, Muscat, Oman, Feb 12-14 (2001)


[16] Antonio Campillo, Patrick Fitzpatrick, Edgar Martinez and R. Pellikaan, “Special Issue Algebraic Coding Theory and Applications”, Journal of Symbolic Computation, July 2010.

[17] R. Togneri, C. DeSilva, “Fundamentals of Information Theory and Coding Design”, Chapman & Hall/CRC, 2006.

[18] San Ling and Chaoping Xing, “Coding Theory”, Cambridge University Press, 2004.


Recovery Method for Disasters of Network Servers by Using POX controllerin Software defined Networks.

1 2 3 4 Asis Jamal , Sarah Javed , Arslan Akram, and Shahzaib Jamal

1Department of Computer Science, Lahore Garrison University, Lahore, PakistanEmail: [email protected], [email protected], [email protected], [email protected]

1. INTRODUCTION

The recovery of disaster is more complex because the network system are becoming more complex to maintain and to control the network system manually. The maintenance of network system is difficult because of increasing growth of network infrastructure so the disaster management is more complex now a days. We can solve the problem by going on that node of network like vender specification solutions and also by debugging that error. For this network complexities we are going to use the technology

named as software defined networking. So SDN is a new and innovative approach to provides the control and flexibility to the network and also used for the managing , building and designing the network infrastructure. In this paper we are going to recover our servers that effected by disaster and the disasters like : traffic engineering, rerouting of packets, device failures and also the link failure. So specifically in this paper we proposed an approach for disaster recovery using software defined networks. SDN provides more programmability by separating the control plane from data plane

Abstract:

The devices we used to automation made up of electric bunch called IoT devices so the more complex task is to manage them effectively. If the devices cannot connect or share anything correctly then these devices will be considered as useless. So the diversity of these devices will be increase the chance of survival. When we talk about the disaster the main difference between networks software and hardware needed to be overcome . for this we have to control the data traffic smartly so the software defined networks make this thing possible to give more programmability because in SDN the data plane is separated from control plane. When IoT devices got disconnected because of internet lack at that time these devices have to respond quickly. So for this purpose, software defined networks used to search another path for information transfer just to get that connection back we can say SDN provides reroute based on the routing information and routing flows they have already and they also have better understanding of pathways for communication. So in this paper our main focus is on this problem that will occur because of disaster and we intend to recover server and also multiple servers from this disaster of link failure , traffic engineering , power outage and rerouting of packets. For this we proposed a systematic approach for recovering these servers from disaster by using software defined networks. The separation of control plane from data plane provides programmability and also make the system flexible for getting back the connection soon. So for this recovery from disaster we are going to use OpenFlow protocol used by SDN and we using Mininet to implementation. The controller will be POX and also we are using Lipsflow mapping for disaster management and recovery.

Keywords: LISP, Locator/ID separation Protocol, SDN, software defined network, Open Flow, load balancing and latency)


Jamal et al LGURJCSIT 2019



45


and also it increases the flexibility and through this SDN can provides disaster recovery very fast[1]. This technology called SDN requires methods for the communication between data plane and control plane and that mechanism or method called OpenFlow protocol which is used by software defined networking for packet matching phenomena. We used POX controller which is written in python and with this controller we used LISP flow mapping for packet mapping and the LISP that will provides the flexible map and framework for the network application. For this purpose the emulator we used is mininet for creating the virtual network[2].

2. RELATED WORK

In network design and management Software defined network gives a different approach[3]. There is static nature of conventional networking even there is small changes in the condition of network that would at high cost of reconfiguring the switches that will be at large number and also routers and other resources of network. Shiaeles et al. (2018) described FHSD solution the FHSD stands for "Fuzzy hybrid spoofing detection". It is a multi-layered spoofing identification system. It uses MAC address, counts hops and web client[4]. It uses the empirical rules for detection of malfunctioned traffic and its mitigation. This strategy accompanies its own disadvantages as its solutions features values are stored in files, this comparisons is cumbersome with it comes with the database. Another method is HCF, the HCF stands for Hop Count Filtering. It utilizes the TTL estimation of the source header to recognize the disaster. Dou et al. (2016) filtering technique uses statistical correlation between different attributes is described. This is used both attack situation and the situation when there is no attack. When there is no attack, normal pattern is analyzed using attribute pairs from the transport layers of the network packets. The recurrence of event of these sets would be extricated and used to find the confidence value of the stream. The attributes that exist between these two layers were utilized to decide the authenticity of a packet[5]. During the attack, the same confidence value is used to find whether an incoming packet is valid or not. The procedure utilizes the Confidence score to learn the authenticity of an incoming packet with comparing it to a threshold. If confidence limit is

satisfied then access is granted to this packet. Aroua and Zouari (2015) have introduced architectural approaches to introduce a coordinated detection with response strategy. The authors consider the existing network architectures that are used by most ISP providers where traffic behavior is analyzed, collected and stored in a central server[6]. However, the authors fears system failure when it comes to the central server or gets compromised by an attacker. In their work, they have improved the single point of failure by equally distributing the shared information using the Byzantine hypothesis of byzantine general problem[7]. When an attack is detected, the information is shared and defense system applied.

A. SDN architecture

The unique feature of SDN is the control plane is separated from data plane . Control plane is formatted by a set of controllers which acts as intelligent brain of SDN, while the data plane formed by multiple packet forwarding switches. This separation of control and data plane enables the network to be directly programmable and achieve benefits like, simple network management, improve the utilization of network, and network innovations and all that. For SDN the communication interface is open flow which is considered as single controller to gain the simplicity. When the network scale continually expands a protocol may suffer from scalability and performance issues. As needs be, various multi-controller approaches are then proposed, and luckily they accomplish a typical fundamental design with joint endeavors. SDN design comprises of three layers: data plane, control plane, and application plane. The data plane is made out of bundle sending switches that are overseen by controllers through southbound application programming interfaces (APIs)[8]. The controllers are associated with the application plane by means of northbound APIs to encourage organize control and system administrations. The main idea of Software Defined Networking (SDN) is the partition of control and information planes. With SDN, the generally circulated control plane of system components, for example, switches and routers, is coherently unified in a SDN controller SDN controller has a worldwide perspective of the system and can settle on better steering choices in view of the present condition of the system, for example interface use one of the hubs. Along


these lines, SDN empowers more proficient system control and administration. In spite of the fact that the control plane is legitimately incorporated, in excess of one controller may be required for adaptability and dependability . High accessibility and low control plane inactivity are important to ensure the information plane execution, which is particularly essential for mission basic applications. The SDN engineering can be separated into six sections, and each part is clarified in detail as takes after: (1) Management plane: It incorporates organizing applications like directing, observing, stack adjusting, and firewalls[9]. Administration plane is in charge of characterizing standards and arrangements. A few creators utilize the term application plane rather than administrat ion plane. (2) . Northbound interface: A northbound interface offers help for correspondence between administration plane and control plane. It gives low level directions to the southbound interface. It is otherwise called administration to control plane interface. Up till now, no standard conventions have been characterized for the northbound interface. (3). Control plane: It is in charge of programming the sending gadgets. It in this way goes about as the mind of the system. Concentrated controllers dwells on this plane[10]. The controller has the total worldwide perspective of the systems . It is otherwise called controller plane. The arrangement of controllers oversees control plane or controller plane. (4) East west interface convention: East West convention is utilized to deal with the correspondence among various SDN controllers[11].

Fig 1. Architecture of software defined network.

B. OpenFlow

OpenFlow Is the protocol of software d e f i n e d n e t w o r k s t h a t p r o v i d e s t h e communication between controller and network device in the architecture of SDN. It was proposed to empower analysts to test new thoughts in a generation domain. OpenFlow gives a particular to relocate the control rationale from a switch into the controller. It likewise delivers a convention for the correspondence between the controller and the switches[12].

C. Pox controller:

Pox is the open source platform for development which is python based software defined networking application like OpenFlow and the other controllers. POX is the tool which enables the development and prototyping rapidly and this is commonly used platform than NOX which is java based platform for results getting[13].

Figure 2.The architecture of POX controller of software defined networks.

D. LISP overview

Another protocol which we are going to use is LISP stands for location id separation protocol. It is the protocol which provides the separation between host locator and host identity w h i c h w i l l d o n e b y c r e a t i n g t w o namespace[14]s. The two namespaces like end point identifier and the host locator so the LISP provides the mapping between virtual IP address and physical IP address . these two namespaces named as EID and RLOC [15].


E. Service of LISP flow mapping:

LISP flow mapping provides the system for mapping. This will have two entities named as map resolver and map server. This will provides the mapping of virtual IP address and physical address[16] . Mapping data can also include a variety of routing policies includes disaster recovery, traffic engineering and load balancing[17].

3. PROPOSED METHDOLOGY

Software defined network(SDN) use the controller named as POX and LISP flow mapping for the recovery of disaster in servers. For effective recovery from disaster using this technology SDN and also the lisp nodes that are mobile nodes and can installed in client and server nodes for the connectivity between server and client and for the stream less connectivity. Software defined networking (SDN) separates the data and control planes, removes the control plane from network hardware and implemented from software instead, which enables programmatic interface and, it increases flexibility of managing network. OpenFlow Is the protocol of software defined networks that provides the communication between controller and network device in the architecture of SDN. It was proposed to empower analysts to test new thoughts in a generation domain. OpenFlow gives a particular to relocate the control rationale from a switch into the controller. It likewise delivers a convention for the correspondence between the controller and the switches. This protocol is the combination of multiple components called plugins , pluggable controllers and also the applications. Locator/identifier separation protocol (lisp) is another protocol which we are going to use is LISP stands for location id separation protocol. It is the protocol which provides the separation between host locator and host identity which will done by creating two namespaces. The two namespaces like end point identifier and the host locator so the LISP provides the mapping between virtual IP address and physical IP address . these two namespaces named as EID and RLOC.

Fig 3. Architecture of proposed system for effective recovery of disaster by using SDN.

The LISP protocol have map server which is an open source environment for development , building and also enable the internet applications in it. The map server authenticate the EID-ROLC mappings by adding them to database. Here is also the command of map resolver which accepts the packets from ingress tunnel router(IRT) then solves the mapping of EID to ROLC by adding them to database. So we are going to resolve this problem or you can say to do effective recovery from disaster on server by using SDN. Because SDN have separated the control plane from data plane which will gives the more flexibility and programmability .control plane is the layer in networks that is responsible for the flow control which have the functions of management and configuration. Data plane is also called as forwarding plane which carries the data packets and you can say it carries the requests. The data plane is responsible for data transfer between clients, handling multiple conversations using various protocols, and manages communication with remote hosts. Data plane packets travels via routers, rather than to or from the. LISP mob is an application which can change their network attachment point without losing connection between host. Lisp mobile node: Lisp mobile node typically sends and receives LISP encapsulated packets. It uses the two name spaces endpoint identifier (EID) to name hosts in networks and routing locators (RLOCs) to locate a node.

4. IMPLEMENTATION

In this paper we have divided the implementation into four parts named as: SDN controller configuration(POX controller), Map registration, server authorization and the priority based checking of server.The flow chart for the recovery of disaster is discussed below in figure


4. and the flow of the implementation is:

Fig4: Flow chart for proposed methodology of disaster recovery of server system

First we have to run the controller or you can say configure the SDN controller and in this paper we consider the POX controller for supporting the lisp flow mapping. Then we install the mobility nodes on the client and server end for the configuration of network. After that we define the mapping system and their key. Key means that the information and registration are stored in flow table and these entries will be identified by a key value generated. And the mapping we discussed is used for recovery as because it will map the physical address and the virtual address that will done by using LISP service. In next step of methodology we setup an environment of virtual switches just to note the fail over between the server 1 and the second server. By using northbound API and southbound API we create the bridge between switches and controller to assign the priority to server and to check the connectivity between server and client to see the failover between server1 and server2 by using mininet. After that the data access object used to separate the database from the user, it will create the bridge between map resolver and the map server this will be done by API’s. the main objective of the DAO is to access the database without knowing the implementation logic. So then the map server is used to registering and adding the server and making key and mapping system. A map resolver is used to processing a

query and to receive a query to process that will received by client and server.

Map registering:

For the map registering the first step is to fetch the end point identifier EID and then store it in locator list or array. After that request for the message of map register. The notify message will received by server to map register after that add up the EID record to the database and then checking for the authentication of data which present. For this the condition is if the key_id matches with server key_id then send the notify message registering to server otherwise send the acknowledgement to server to notify that this no authentication of data have done.

Fig5. Flowchart for the map registering in disaster recovery.

Server authentication:

After the map registering there is need to authenticate the server so we have done server authentication for this the first step is to map server should be authenticated then get the key of authentication from the map resolver and then map server will sends iterate mask to map resolver and if the map resolve sends notify message then add key to authentication key if its not getting any notify message then remove authentication key from map server.


Fig 6. Figure or flow of steps for server authentication.

Servers priority checking:

The pseudocode for the priority checking this discussed bellow first, client send request to server by SDN controller named as POX which follows the OpenFlow protocol. Then if the client key_id matches with the server key_id it will send message that controller connects with server. If its not connects then will show message that the controller disconnects the client by sending message that key is not matched with map server key. After that or in next step disaster occurs at server1 and it crashes. Then controller will check the priority by map resolver. The condition is set if the server1 priority is less than priority of server2 then it will migrate towards the server2 while if its not then it will check for the next server which one having higher priority. Flow chart for priority checking: In this flow chart figure 8 we intended to explain checking priorities server by SDN controller. The client sends request to server via controller, the controller checking key_id if key_id matches the controller connects to server for service otherwise controller send notify message called key not matches with map server key_id . the client get service on same time server gets disaster the controller searching for another server with higher priority value in this server2 having higher priority so it’s connects to server2 if priority lower so it searching for another server.

5. EXPERIMENTAL RESULTS

In this research there are two parameters Load balancing and latency. So, the Latency is

the delay from input into a system to desired outcome; the term is understood slightly differently in various contexts and latency issues also vary from one system to another. Latency greatly affects how usable and enjoyable electronic and mechanical devices as well as communications are People connecting from distances to these live events can be seen to have to wait for responses. This latency is the wait time introduced by the signal travelling the geographical distance as well as over the various pieces of communications equipment. Even fiber optics are limited by more than just the speed of light, as the refractive index of the cable and all repeaters or amplifiers along their length introduce delays. And the second one is load balancing refers to efficiently distributing incoming network traffic across a group of


Fig7.Architecture for priority checking

backend servers, also known as a server farm or server pool. To cost-effectively scale to meet these high volumes, modern computing best practice generally requires adding more servers.

Now the second parameter is latency or delay at same requirements when each node forward five packets per second and ten packets per second so the result is like and the simulation environment is:

6. CONDLUSION

In this paper we have implements the disaster recovery of server by using the technology software defined networks and the LISP the flow mapping service over UBUNTU operating system successfully. We have implemented the map registering by EID , the mapping service of the requests and by priority checking which have done by SDN controller named as POX controller which is open source the python based development environment. In future we can extend this work to deploy lisp flow mapping In android and also in windows. So we can extend the platform. And we can also see the disaster in case of load balancing by end to end communication.

References

[1] K. Nguyen, Q. T. Minh, and S. Yamada, "A software-defined networking approach for disaster-resilient WANs," in 2013 22nd International Conference on Computer Communication and Networks (ICCCN), 2013, pp. 1-5.

[2] S. Mehraghdam, M. Keller, and H. Karl, "Specifying and placing chains of virtual network functions," in 2014 IEEE 3rd International Conference on Cloud Networking (CloudNet), 2014, pp. 7-13.

[3] W. H. Muragaa, K. Seman, and M. F. Marhusin, "A pox controller module to collect web traffic statistics in SDN environment," World Academy of Science, Engineering and Technology, International Journal of Computer, Electrical, Automation, Control and Information Engineering, vol. 10, pp. 2002-2007, 2016.

[4] F. Bannour, S. Souihi, and A. Mellouk, "Distributed SDN control: Survey, taxonomy, and challenges," IEEE Communications Surveys & Tutorials, vol. 20, pp. 333-354, 2018.

[5] M. Ziaullah, P. Shetty, and S. Kamal, "Image feature based authentication and digital signature for wireless data transmission," in


2016 International Conference on Computer Communication and Informatics (ICCCI), 2016, pp. 1-4.

[6] R. T. Baum, "IP based securi ty applications using location, port and/or device identifier information," ed: Google Patents, 2011.

[7] M. Kamruzzaman, N. I. Sarkar, J. Gutierrez, and S. K. Ray, "A study of IoT-based p o s t - d i s a s t e r m a n a g e m e n t , " i n 2 0 1 7 International Conference on Information Networking (ICOIN), 2017, pp. 406-410.

[8] R. Mijumbi, J. Serrat, J.-L. Gorricho, N. Bouten, F. De Turck, and R. Boutaba, "Network function virtualization: State-of-the-art and research challenges," IEEE Communications Surveys & Tutorials, vol. 18, pp. 236-262, 2016.

[9] C. P. Dingman, P. Mahadevan, and J. J. Ordille, "Method and apparatus for supporting communications between a computing device within a network and an external computing device," ed: Google Patents, 2016.

[10] L. R. Dennison, "Software control plane for switches and routers," ed: Google Patents, 2013.

[11] M. Karakus and A. Durresi, "A survey: Control plane scalability issues and approaches in software-defined networking (SDN)," Computer Networks, vol. 112, pp. 279-293, 2017.

[12] J . Wa n g a n d M . L u o , " P a c k e t prioritization in a software-defined network implementing OpenFlow," ed: Google Patents, 2018.

[13] V. Gramoli, G. Jourjon, and O. Mehani, "Disaster-tolerant storage with SDN," in International Conference on Networked Systems, 2015, pp. 293-307.

[14] J. R. Putman, M.-H. Nguyen, T. C. Hanson, and S. Srinivasan, "Communication a p p l i c a t i o n s e r v e r f o r c o n v e r g e d communication services," ed: Google Patents, 2015.

[15] P. Kakade, S. B. Raman, and R. Sharma, "Systems and methods for business impact analysis and disaster recovery," ed: Google Patents, 2019.

[16] S. Azodolmolky, Software defined networking with OpenFlow: Packt Publishing Ltd, 2013.

[17] G. Liu, M. Trotter, Y. Ren, and T. Wood, "Ne ta ly t i c s : C loud- sca le app l i ca t ion performance monitoring with sdn and nfv," in Proceedings of the 17th International Middleware Conference, 2016, p. 8.


Software Testing for cyber Security

Arfa Hassan , Tayyaba Anees, Adnan Khan

1. Introduction

Software testing is the process of software quality assurance which is used to figure out the faults, defects, errors and bugs of t h e d e v e l o p i n g a n d a s w e l l a s t h e underdeveloped software.[1] Experts divide the software testing phases into three categories i.e., functional testing, nonfunctional testing and regression testing. Software testing is constructive to reduce the maintenance cost. And it is also used to determine the security risks of the software before its deployment. The security of any software is an important feature because majority of activities of today’s human is controlled through software[2]. And thus software security barrier can cause the unbearable loss. An ACI (army cyber institute) report shows that innovation in the cloud computation and IOT devices make a global shortage of cyber-security talent in 2017. This gap between cyber security talent and white collar criminals can create the political, personal and business consequences such as panama leaks and wiki leaks. [3]The reason behind these cyber security attacks are that companies show less interest towards cyber security as

sometimes the authorities of these companies have not enough awareness regarding white collar crime and also another major issue is that the different vendors of the devices follows different kind of rules for the cyber security[4]. The most recent incident in this context occurs with Careem ride service in Pakistan. The hackers hack their website and steel all the important information of their customer. The stolen data also includes some personal information like their credit card details etc. The main focus of hackers in such incidents are to steel the banking details, phone numbers and other personal details which they can further use for illegal activities[5]. The process of identifying, analysis and e v a l u a t i o n o f s o f t w a r e b e f o r e t h e implementation of that software is important for its protection and unauthorized handling. The results of the survey 2017 on cyber security breaches arranged by the UK government reveals that main cause of cyber-attacks is that the companies do not care about the cyber security at foundational levels because they think that they are not at risk. Cyber security risk assessment is equally important for the hardware because many hackers also able to access the

Abstract:

Software testing plays a vital role in software security because hackers attack on a system through back channels which they can easily find if there is any error or bug exists in the software. The software security failure can cause the unbearable loss for IT companies and other organizations. Cyber security is another big issue for computer users' personal data as all their information is vulnerable because of easy excess, visibility and availability. Therefore, software testing is also useful to secure the personal information. In this article, cyber security testing based on particle swam optimization algorithm (CST) is proposed for testing of software cyber security testing. CSTPSOA is a PSO base technique which is used to solve the complex multi-level problems and is also used for optimization. In the CST method PSO is used for the optimization of test cases for cyber security testing.

Keywords: Cyber Security, CST, Software Testing, PSO, Quality Assurance


Hassan et al LGURJCSIT 2019



53


hardware component such as Ram, Rom and attached device, for example, USB, external hard drive, printer etc.[5]. Test case generation and evolution is an important task for the SQA professional. This research article CST method is proposed for software testing. CST algorithm used Particle swarm optimization technique for cyber security testing test case optimization. PSO is a greedy approach of artificial intelligent to get best results in less time[6].

Literature Reviews

Researches introduce different kinds of algorithm for the software testing and software security testing. A few of them are discussed below. PSO based greedy algorithm approach is very useful in regression testing. Regression testing looks expensive, but it is really helpful to reduce the maintenance cost of the software. For this purpose the researchers takes the sample set of 5 different softwares and test them line by line through the sample test set[7]. Web base application is main source of business, communication and information sharing so protection of web based application from hackers attack is necessary. Software security testing before implementation of such type of applications can play important role. In [8] article the researcher used the genetic algorithm based approach for web based applications security testing. This proposed method could be able to perform analysis of expected breaches and divide the problem in two categories after analysis.

• Static analysis• Dynamic analysis

This analysis of the software creates the fitness function and check the security issues of the software. Automatic software testing tools can play vital role in software quality assurance. Software quality assurance helps to reduce about 50% of software development cost[9]. The proposed algorithm of this purpose is feed forward neural network with back propagation learning algorithm. Two models are used in this approach. Model 1 takes 2 input variables but in model 2 researchers uses 4 variables[9]. In another article, the researchers work on fault detection techniques. For this purpose, they

apply Rank-to-learn algorithm. To enhance the performance of rank-to-learn technique they also use the back propagation. To evaluate this system they take 10 different cases. For the simulation matlab is used[10]. The basic aim of today’s software industry is to provide high quality secure software to the end users. Software testing techniques can play a vital role in the software quality assurance. The field of software testing have various issues which require efforts on time and cost of testing. For this purpose the researcher conduct the survey on Genetic algorithm techniques. In this article number of techniques are discussed which use genetic algorithm techniques for software testing tools development[11].

Proposed Method

In this article CST method is proposed for software security testing. CST works on the Particle swam optimization. PSO is stochastic optimization technique[12]. PSO can be applied on various problems of different fields. PSO is a population based computational problem solving method[7]. The basic idea of PSO is taken form the flak of swarms[13]. The step of PSO is as following:

• Initial the population• Evaluate the fitness• Find the particle best value • Find the Global best Value • Create the stopping condition.

The mathematical formulation of particle best and global best values is given below:pb(s,i)=argmin[f(Ls(k))] where s є [1,2,3…….n] gb(i)=argmin[f(Li(k))] where i = [ 1 , 2 , 3 … … . n ] where k=[1,2,3,4……i]µ s ( i + 1 ) = ώ µ s ( i + 1 ) + a 1 d 1 ( p b ( s , i ) ) -Ls(i))+a2d2(gb(i)- Ls(i)) 1L s ( i + 1 ) = L s ( i ) + µ s ( i + 1 ) 2Wherepb= personal bestgb= global bests= particle indexn= total number of si = current iteration numberf = fitness function


L = positionµ = velocityώ = weightd1 & d2= random variablesa1 & a2= acceleration coefficients.

whole software in a binary form, then encode it to remove 0 values form the code. After initialization of the test case population, optimization process is start. For this purpose algorithm evaluate the fitness function and then finds the local best or global best value. If the global best value are between the fitness functions value or required number of iterations are complete then stop this process .The complete process of proposed CST method is shown in figure 1.

Simulation and results

For simulation and results Matlab R2017a tool is used. For simulation and results Matlab R2017a tool is used. The proposed system further illustration through an example. Table 1 shows the population and table 2 shows the fitness test.

Tab le 1 : BPSK Tes t case popu la t ion initialization.

Table 2: BPSK Fitness function The cost function of proposed CST based pso method is (Gbesti/Pbesti) .xi


Figure 1: Flow chart of CST

Software security is a sensitive issue so many different techniques are developed by the SQA professionals for software testing. Formal method [14] and model based techniques are one them. NASA[2] introduced formal methods for software testing. Formal method is a way of software testing in which whole software is converted into pieces of some meaningful algorithm[15]. In the model based software testing approach software tests are designed through the UML [16] diagrams of software and generate test case through the standard software[17]. The proposed method of this article is CST (cyber security testing) based on particle swarm optimization algorithm (PSO). For this purpose the first step is to convert the

Figure 2 shows the matlab graph results of proposed PSO based CST system.

Figure 2.PSO with different number of cycles

Conculsion

Cyber security is a sensitive issue and required great attention. But actually this is the most neglected part of SDLC especially in under developed countries like Pakistan. Campiness can easily secure their products by performing proper software security testing on their products. Test case generation is major and the most challenging part of software testing process. Sometime thousands of test cases are generated by the software, which makes the sof tware tes t ing job more tough and complicated. In this article CST algorithm is proposed for test case optimization. The proposed method uses the PSO algorithm for optimization of test cases. The proposed method could also be used to find out the perfect test cases on the basis of fitness test. REFERENCES

[1] J. J. Scarpino, “Web Application Security Testing : an Industry Perspective on How Its Education Is Perceived,” vol. XI, no. 1, pp. 142–153, 2010.

[2] Gu Tian-yang, Shi Yin-sheng, and Fang You-yuan, “Research on Software Security Testing,” Int. J. Comput. Electr. Autom. Control Inf. Eng., vol. 4, no. 9, pp. 1447–1450, 2010.

[3] A. Bendovschi, “Cyber-Attacks – T r e n d s , P a t t e r n s a n d S e c u r i t y Countermeasures,” Procedia Econ. Financ., vol. 28, no. April, pp. 24–31, 2015.

[4] J. Twist, “ACI Threat Trends and Predictions 2017 Report,” 2017.

[5] B e c k y M e t i v i e r, “ 6 S t e p s t o a Cybersecurity Risk Assessment,” sage data s e c u r i t y, 2 0 1 7 . [ O n l i n e ] . Av a i l a b l e : https://www.sagedatasecurity.com/blog/6-steps-to-a-cybersecurity-risk-assessment. [Accessed: 26-Mar-2018].

[6] D. Palupi Rini, S. Mariyam Shamsuddin, and S. Sophiyati Yuhaniz, “Particle Swarm Optimizat ion: Technique, System and Challenges,” Int. J. Comput. Appl., vol. 14, no. 1, pp. 19–27, 2011.

[7] U. Jafri, H. Sadia, and J. Ahmad, “PSO based Optimized Software Testing Technique,” vol. 13, no. 2, 2017.

[8] A. Avancini and F. B. Kessler, “Security Testing of Web Applications : A Research Plan,” no. line 1, pp. 1491–1494, 2012.

[9] Q. P. Rana, “Model for Software Testing and Quality Assessment using ANN Approach,” vol. 14, no. 1, 2017.

[10] P. K. Sidhu, “Evaluate and Propose Fault Detection Technique from Test Cases in Software Testing,” vol. 5, no. 3, pp. 9–14, 2017.

[11] C. Sharma, S. Sabharwal, and R. Sibal, “A Survey on Software Testing Techniques using Genetic Algorithm,” vol. 10, no. 1, pp. 381–393, 2013.

[12] A. June, “Research Paper on Optimized Utilization of Resources Using PSO and Improved Particle Swarm Optimization ( IPSO ) Algorithms in Cloud Computing,” vol. 2, no. 2, pp. 499–505, 2014.

[13] Eberhart and Yuhui Shi, “Particle swarm optimization: developments, applications and resources,” Proc. 2001 Congr. Evol. Comput. (IEEE Cat. No.01TH8546), vol. 1, no. February, pp. 81–86, 2015.


[14] C. M. May, “Formal Methods for IT Security,” pp. 1–44, 2007.

[15] A. Bertolino, “Software Testing Research: Achievements, Challenges, Dreams,” Futur. Softw. Eng. (FOSE ’07), no. September, pp. 85–103, 2007.

[16] I. Schieferdecker, J. Grossmann, and M. Schneider, “Model-Based Security Testing,” Electron. Proc. Theor. Comput. Sci., vol. 80, no. Mbt, pp. 1–12, 2012.

[17] A. Bertolino, “Software Testing Research and Practice,” Abstr. State Mach. 2003, pp. 1–21, 2003.


Date post:	23-Sep-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

An Intelligent and Secure Communication of AIoT enabled...

Documents