Applications of Cognitive Radio Networks Recent Advances and Future Diections

8/19/2019 Applications of Cognitive Radio Networks Recent Advances and Future Diections

1/70

International Journal of Distributed Sensor Networks

Applications of Cognitive Radio

Networks: Recent Advances andFuture Directions

Guest Editors: Sung W. Kim, Miao Pan, Gyanendra Prasad Joshi, Orhan Gazi,

Jianhua He, and Marceau Coupechoux


2/70

Applications of Cognitive Radio Networks:

Recent Advances and Future Directions


3/70

International Journal o Distributed Sensor Networks

Applications of Cognitive Radio Networks:

Recent Advances and Future Directions

Guest Editors: Sung W. Kim, Miao Pan,

Gyanendra Prasad Joshi, Orhan Gazi, Jianhua He,

and Marceau Coupechoux


4/70

Copyright © Hindawi Publishing Corporation. All rights reserved.

Tis is a special issue published in “International Journal o Distributed Sensor Networks.” All articles are open access articles distributedunder the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro-

vided the original work is properly cited.


5/70

Editorial Board

Jemal H. Abawajy, AustraliaMiguel Acevedo, USACristina Alcaraz, SpainAna Alejos, SpainMohammod Ali, USAGiuseppe Amato, Italy Habib M. Ammari, USAMichele Amoretti, Italy Christos Anagnostopoulos, UKLi-Minn Ang, AustraliaNabil Aou, UKFrancesco Archetti, Italy Masoud Ardakani, CanadaMiguel Ardid, SpainMuhammad Asim, UKSteano Avallone, Italy Jose L. Ayala, SpainJavier Bajo, SpainN. Balakrishnan, IndiaPrabir Barooah, USA

Federico Barrero, SpainPaolo Barsocchi, Italy Paolo Bellavista, Italy Olivier Berder, FranceRoc Berenguer, SpainJuan A. Besada, SpainGennaro Boggia, Italy Alessandro Bogliolo, Italy Eleonora Borgia, Italy Janos Botzheim, JapanFarid Boussaid, AustraliaArnold K. Bregt, NetherlandsRichard R. Brooks, USAed Brown, USADavide Brunelli, Italy James Brusey, UKCarlos . Calaate, Spainiziana Calamoneri, Italy José Camacho, SpainJuan C. Cano, SpainXianghui Cao, USAJoão Paulo Carmo, BrazilRoberto Casas, SpainLuca Catarinucci, Italy Michelangelo Ceci, Italy

Yao-Jen Chang, aiwanNaveen Chilamkurti, AustraliaWook Choi, Republic o KoreaH. Choo, Republic o KoreaKim-Kwang R. Choo, AustraliaChengu Chou, aiwanMashrur A. Chowdhury, USAae-Sun Chung, Republic o KoreaMarcello Cinque, Italy Sesh Commuri, USAMauro Conti, Italy Alredo Cuzzocrea, Italy Donatella Darsena, Italy Dinesh Datla, USAAmitava Datta, AustraliaIyad Dayoub, FranceDanilo De Donno, Italy Luca De Nardis, Italy Floriano De Rango, Italy Paula de oledo, Spain

Marco Di Felice, Italy Salvatore Disteano, Italy Longjun Dong, ChinaNicola Dragoni, Denmark George P. Efhymoglou, GreeceFrank Ehlers, Italy Melike Erol-Kantarci, CanadaFarid Farahmand, USAMichael Farmer, USAFlorentino Fdez-Riverola, SpainGianluigi Ferrari, Italy Silvia Ferrari, USAGiancarlo Fortino, Italy Luca Foschini, Italy Jean Y. Fourniols, FranceDavid Galindo, SpainEnnio Gambi, Italy Weihua Gao, USAA.-J. García-Sánchez, SpainPreetam Ghosh, USAAthanasios Gkelias, UKIqbal Gondal, AustraliaFrancesco Grimaccia, Italy Jayavardhana Gubbi, AustraliaSong Guo, Japan

Andrei Gurtov, FinlandMohamed A. Haleem, USAKijun Han, Republic o KoreaQi Han, USAZdenek Hanzalek, Czech RepublicShinsuke Hara, JapanWenbo He, CanadaPaul Honeine, FranceFeng Hong, ChinaChin-ser Huang, USAHaiping Huang, ChinaXinming Huang, USAJose I. Moreno, SpainMohamed Ibnkahla, CanadaSyed K. Islam, USALillykutty Jacob, IndiaWon-Suk Jang, Republic o KoreaAntonio J. Jara, SwitzerlandShengming Jiang, ChinaYingtao Jiang, USA

Ning Jin, ChinaRaja Jurdak, AustraliaKonstantinos Kalpakis, USAIbrahim Kamel, UAEJoarder Kamruzzaman, AustraliaRajgopal Kannan, USAJohannes M. Karlsson, SwedenGour C. Karmakar, AustraliaMarcos D. Katz, FinlandJamil Y. Khan, AustraliaSheri Khattab, EgyptHyungshin Kim, Republic o KoreaSungsuk Kim, Republic o KoreaAndreas König, Germany Gurhan Kucuk, urkey Sandeep S. Kumar, NetherlandsJuan A. L. Riquelme, SpainYee Wei Law, AustraliaAntonio Lazaro, SpainDidier Le Ruyet, FranceJoo-Ho Lee, JapanSeokcheon Lee, USAYong Lee, USASteano Lenzi, Italy Pierre Leone, Switzerland


6/70

Shancang Li, UKShuai Li, USAQilian Liang, USAWeia Liang, AustraliaYao Liang, USAI-En Liao, aiwanJiun-Jian Liaw, aiwanAlvin S. Lim, USAAntonio Liotta, NetherlandsDonggang Liu, USAHai Liu, Hong KongYonghe Liu, USALeonardo Lizzi, FranceJaime Lloret, SpainKenneth J. Loh, USAJuan Carlos López, SpainManel López, SpainPascal Lorenz, FranceJun Luo, SingaporeMichele Magno, Italy Sabato Manredi, Italy Athanassios Manikas, UKPietro Manzoni, Spain

Álvaro Marco, SpainJose R. Martinez-de Dios, SpainAhmed Mehaoua, FranceNirvana Meratnia, NetherlandsChristian Micheloni, Italy Lyudmila Mihaylova, UKPaul Mitchell, UKMihael Mohorcic, SloveniaJosé Molina, SpainAntonella Molinaro, Italy Salvatore Morgera, USAKazuo Mori, JapanLeonardo Mostarda, Italy V. Muthukkumarasamy, AustraliaKamesh Namuduri, USAAmiya Nayak, CanadaGeorge Nikolakopoulos, SwedenAlessandro Nordio, Italy

Michael J. O’Grady, IrelandGregory O’Hare, IrelandGiacomo Oliveri, Italy Saeed Olyaee, IranLuis Orozco-Barbosa, SpainSuat Ozdemir, urkey Vincenzo Paciello, Italy Sangheon Pack, Republic o KoreaMarimuthu Palaniswami, AustraliaMeng-Shiuan Pan, aiwanSeung-Jong Park, USAMiguel A. Patricio, SpainLuigi Patrono, Italy Rosa A. Perez-Herrera, SpainPedro Peris-Lopez, SpainJanez Perš, SloveniaDirk Pesch, IrelandShashi Phoha, USARobert Plana, FranceCarlos Pomalaza-Ráez, FinlandNeeli R. Prasad, Denmark Antonio Puliato, Italy Hairong Qi, USA

Meikang Qiu, USAVeselin Rakocevic, UKNageswara S.V. Rao, USALuca Reggiani, Italy Eric Renault, FranceJoel Rodrigues, PortugalPedro P. Rodrigues, PortugalLuis Ruiz-Garcia, SpainMohamed Saad, UAESteano Savazzi, Italy Marco Scarpa, Italy Arunabha Sen, USAOlivier Sentieys, FranceSalvatore Serrano, Italy Zhong Shen, ChinaChin-Shiuh Shieh, aiwanMinho Shin, Republic o KoreaPietro Siciliano, Italy

Olli Silven, FinlandHichem Snoussi, FranceGuangming Song, ChinaAntonino Staiano, Italy Muhammad A. ahir, PakistanJindong an, USAShaojie ang, USALuciano arricone, Italy Kerry aylor, AustraliaSameer S. ilak, USAChuan-Kang ing, aiwanSergio oral, SpainVicente raver, SpainIoan udosa, Italy Anthony zes, GreeceBernard Uguen, FranceFrancisco Vasques, PortugalKhan A. Wahid, CanadaAgustinus B. Waluyo, AustraliaHonggang Wang, USAJianxin Wang, ChinaJu Wang, USAYu Wang, USA

Tomas Wettergren, USARan Wolff, IsraelChase Wu, USANa Xia, ChinaQin Xin, Faroe IslandsChun J. Xue, Hong KongYuan Xue, USAGeng Yang, ChinaTeodore Zahariadis, GreeceMiguel A. Zamora, SpainHongke Zhang, ChinaXing Zhang, ChinaJiliang Zhou, Chinaing L. Zhu, USAXiaojun Zhu, ChinaYieng Zhu, USADaniele Zonta, Italy


7/70

Contents

Applications of Cognitive Radio Networks: Recent Advances and Future Directions

Sung W. Kim, Miao Pan, Gyanendra Prasad Joshi, Orhan Gazi, Jianhua He, and Marceau CoupechouxVolume , Article ID , pages

An Analytical Approach to Opportunistic Transmission under Rayleigh Fading Channels

Yousa Bin Zikria, Sung Won Kim, Heejung Yu, and Seung Yeob NamVolume , Article ID , pages

A Cognitive-Radio-Based Method for Improving Availability in Body Sensor Networks

Olga León, Juan Hernández-Serrano, Carles Garrigues, and Helena Rià-PousVolume , Article ID , pages

Energy-Efficient Layered Video Multicast over OFDM-Based Cognitive Radio Systems

Wenjun Xu, Shengyu Li, Yue Xu, Zhiyong Feng, and Jiaru LinVolume , Article ID , pages

A Cross-Layer-Based Routing Protocol for Ad Hoc Cognitive Radio Networks

Gyanendra Prasad Joshi, Seung Yeob Nam, Chang-Su Kim, and Sung Won KimVolume , Article ID , pages

Sensor Virtualization Module: Virtualizing IoT Devices on Mobile Smartphones for Effective Sensor

Data Management

JeongGil Ko, Byung-Bog Lee, Kyesun Lee, Sang Gi Hong, Naesoo Kim, and Jeongyeup Paek Volume , Article ID , pages

Convergence Research Directions in Cognitive Sensor Networks for Elderly Housing DesignShinil Suh, Byung-Seo Kim, and Jae Hee ChungVolume , Article ID , pages


8/70

Editorial Applications of Cognitive Radio Networks:Recent Advances and Future Directions

Sung W. Kim,1 Miao Pan,2 Gyanendra Prasad Joshi,1 Orhan Gazi,3

Jianhua He,4 and Marceau Coupechoux 5

Department of Information and Communication Engineering, Yeungnam University, Gyeongsan , Republic of KoreaDepartment of Computer Science, Texas Southern University, Houston, TX , USADepartment of Electric and Communication Engineering, Cankaya University, Ankara, Turkey Department of Electrical, Electronic and Power Engineering, Aston University, Birmingham B ET, UK Department of Computer and Network Science, Telecom ParisTech, Paris, France

Correspondence should be addressed to Gyanendra Prasad Joshi; [email protected]

Received December ; Accepted December

Copyright © Sung W. Kim et al. Tis is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Contemporary wireless networks must meet the ever increas-ing bandwidth requirements to assure the quality o service(QoS) to the end users. Cognitive radio (CR) technology with efficient electromagnetic spectrum management canachieve increased bandwidth beyond its traditional limits.Te innovative spectrum management by CRNs allows orusage o incumbent spectrum band by unlicensed (cogni-tive) users possibly without interering with the incumbentusers. CR network is an intelligent and adaptive wirelesscommunications system in which CR devices learn rom itssurroundings anddecide the operation based on the learning.CR devices are inevitably so intelligent that they can dynami-cally choose carrier requency, bandwidth, transmission rate,

transmission power, and so orth.Terearemanyemerging CR networks applications based

on CR technologies. Tis special issue is ocused on present-ing state-o-the-art research results on the application o CR networks. Tis is targeted or the innovative and productivediscussion on the recent advancement in the application o CR networks and uture directions.

Te article “Convergence Research Directions in Cog-nitive Sensor Networks or Elderly Housing Design” by S.Suh et al. is about the application o CR sensor network. Itdenes smart home and surveys CR sensor network- (CRSN-) based systems or elderly housing. Tis article proposesresearch directions or the elderly smart home services based

on CRSN. Particularly, the article is ocused on adoptingCRSN technologies to cope with dense sensors environmentand heterogeneous network environment. It also discussescustomizing sensors/networks classication correlated withthe elderly types, and converging sensor network technolo-gies with architectural technologies.

A lightweight and robust mechanism that appropriately secures the channel selection process is presented in thearticle “A Cognitive-Radio-Based Method or ImprovingAvailability in Body Sensor Networks” by O. León et al. Inthis article, authors describe a new network paradigm knownas cognitive body sensor networks (CBSNs). In the body area networks, seamless connectivity is crucial and must be

guaranteed. Connectivity losses during emergency situationsmay prevent a patient rom immediately receiving medicalassistance and may end up in catastrophic results. Tis articlediscusses how to prevent CBSNs rom the specic attacksby securing the sensing process. Te proposed methodrelies on cryptographic primitives that require a minimumamount o memory and low energy consumption, thus beingmore suited or devices with limited resources. It offersauthentication and encryption o control data shared by thesensors in the CBSN to agree on a given channel.

An energy-efficient layered video multicast (LVM) trans-mission over OFDM-based CR systems or “bandwidth-hungry” video services is presented in the article entitled

Hindawi Publishing CorporationInternational Journal of Distributed Sensor NetworksVolume 2016, Article ID 4964068, 2 pageshttp://dx.doi.org/10.1155/2016/4964068

http://dx.doi.org/10.1155/2016/4964068http://dx.doi.org/10.1155/2016/4964068


9/70


“Energy-Efficient Layered Video Multicast over OFDM-Based Cognitive Radio Systems” by W. Xu et al. Tis articleproposes an energy utility- (EU-) based power allocationalgorithm by jointly employing ractional programming andsubgradient method. Te novel perormance metric EU isproposed to measure the sum o achieved quality o recon-

structed video at all subscribers when unit transmit poweris consumed. Te objective is to maximize the system EUby jointly optimizing the intersession/interlayer subcarrierassignment and subsequent power allocation. o achieve theobjective, it perorms subcarrier assignment or base layerand enhancement layers using greedy algorithm and thenpresents an optimal power allocation algorithm to maximizethe achievable EU using ractional programming.

In the article “A Cross-Layer-Based Routing Protocol orAd Hoc Cognitive Radio Networks” by G. P. Joshi et al.,the authors propose a cross-layer-based routing protocol ormobile ad hoc CR networks. Te motivation or this paperis that rerouting is expensive in terms o energy, delay, and

throughput. Tus, it is better to select a route in such a way that requires less channel switching. Tis paper examines theexpectation o channel switching in the range o scenariosand proposes a novel route selection method to mitigate therequent channel switching. Because excessive workload ona particular node causes network partitioning and inducesrepeated rerouting, the proposed protocol distributes therouting overheads among cognitive users in the network andprolongs the network lietime. Tis protocol incorporatespower awareness and spectrum inormation with a cross-layer approach.

Te Sensor Virtualization Module (SVM) is proposed inthe article “Sensor Virtualization Module: Virtualizing IoDevices on Mobile Smartphones or Effective Sensor DataManagement” by J. Ko et al. Tere are limited Io resource-utilizing applications due to the traditionalstovepipe sofwarearchitecture, where the vendors provide supporting sofwareon an end-to-end basis. Te proposed SVM in this articleprovides a sofware abstraction or external Io objects andallows applications to easily utilize various Io resourcesthrough open APIs. It also presents the applications with acommon virtualized environment where external Io devicescan be easily accessed rom and via mobile computingplatorms.

In the article “An Analytical Approach to Opportunisticransmission under Rayleigh Fading Channels” by Y. B.Zikria et al., the authors present the effectiveness o theopportunistic transmission in terms o reliability and delay o transmission analytically. A xed-distance-based statisticalmodel is proposed or multihop and opportunistic trans-mission or CRSNs. Also, a unique generic Markov chainmodel is proposed to show the stability o opportunistictransmission.

Acknowledgments

Tis work was supported by the Yeungnam University Research Grant. Finally, we thank all the authors who sub-mitted their work to this special issue and many experts whoparticipated in the review process. We expect that this issue

will certainly help researchers, academicians, practitioners,and industrialists to realize the recent advances and also may help them to work in the uture directions.

Sung W. Kim Miao Pan

Gyanendra Prasad JoshiOrhan Gazi Jianhua He

Marceau Coupechoux


10/70

Research Article An Analytical Approach to Opportunistic Transmission underRayleigh Fading Channels

Yousaf Bin Zikria, Sung Won Kim, Heejung Yu, and Seung Yeob Nam

Department of Information and Communication Engineering, Yeungnam University, - Dae-dong, Gyeongsan-Si,Kyongsan, Gyeongsangbuk-do , Republic of Korea

Correspondence should be addressed to Sung Won Kim; [email protected]

Received August ; Accepted November

Academic Editor: Lillykutty Jacob

Copyright © Yousa Bin Zikria et al. Tis is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In cognitive radio sensor networks, the routing methods including multiple relays have been extensively studied to achieve higherthroughput and lower end-to-end delay. As one o innovative approaches, the opportunistic routing scheme was proposed. Inthis paper, the effectiveness o the opportunistic transmission in terms o reliability and delay o transmission is veried with ananalytical way. For the analysis, we establish the probabilistic model with respect to distance and the number o relay nodes underthe Rayleigh ading channels including path loss effects. Under this model, we develop a generic Markov chain model to obtain theanalyticalresults and veriy the effectiveness o the statistical analysis. Te results show thatan opportunistic transmission approach

is betterthan traditional multihop transmissions in terms o successuldata delivery with ewer transmissions. Consequently, it canprovide an energy efficient transmission mechanism or cognitive radio sensor networks.

1. Introduction

Reliable data delivery with the ewest hops, keeping end-to-end delay and overhead minimized, is always a prime ocusin cognitive radio sensor networks research that results inincreased throughput. Moreover, the effectiveness o cogni-tive radio sensor networks is dependent on the developmento the effective and energy efficient protocols. Te key idea in

opportunistic routing is to exploit the probability o reachingthe arthest node in one transmission. I we can transmit apacketsuccessully, directly or with the ewesthops,even withlow probability, we can drastically improve throughput andreduce end-to-end delay. Te key challenge o this research isto analyze opportunistic transmission (O) statistically andshow that it is better in terms o successul transmissionsand requires ewer transmissions, compared to traditionalmultihop transmissions, under the assumption that end-to-end distance is known.

Te cognitive radio sensor networks are powered by niteenergy resources. Recent trends in cognitive radio sensornetworks [] and introduction o wireless multimedia sensor

networks [] highlight the importance o energy consump-tion. Terore, more research is inclined to increase thecognitive radio sensor network lietime []. ransmission o packets in multihop wireless networks poses a great challengebecause o unreliability and inherent intererence o wirelesslinks []. Wireless multihop networks [–] encompassmobile or stationary stations interconnected via an ad hocmultihop path. Each node operates not only as a host but

also as a router and orwards packets on behalo other nodesthat may not be within direct radio range o the destinations.Among recent advances, opportunistic routing has appearedas an appealing multihop routing method, which gives highthroughput in dynamic wireless environments.

Opportunistic routing (OR) [–] takes advantage o thespatial diversity and broadcast nature o wireless networks tocombat time-varying links by involving multiple neighboringnodes, also known as orwarding candidates, or each packettransmission []. Adopting a different philosophy in routeselection, O chooses the closest node to the destination toorward a packet out o the set o nodes that actually receivedprevious packets. Tis results in high expected progress per




11/70


transmission. Te exibility o O enables agile adaptation inast-changing wireless environments, which are particularly suitable or serving up high-rate and delay-sensitive interac-tive traffic [].Extremely opportunistic routing (ExOR) inte-grates routing and medium access control (MAC) protocols.It improves throughput by selecting long-range, but lossy,

links. It is designed or batch orwarding. Te source nodeincludes the list o orwarders in a packet, based on expectedtransmission distance rom the destination. All packets arebroadcast. Each packet contains a BIMAP option, whichmarks the successully received packet by the receiver orhigher priority nodes. However, this protocol reduces spatialreuse as it is globally synchronized, and there are duplicatetransmissions as well.

Opportunistic any-path orwarding (OAPF) [] over-comes the problem o ExOR choosing low-quality routes.It introduces an expected path-count metric. Tis approachrecursively calculates the near optimal orwarder set at eachorwarder. However, this approach incurs high computa-tional overhead. MAC-independent opportunistic routingand encoding (MORE) [] integrates a network coding OR to enhance ExOR. Te core idea is to avoid any duplicationo data. It uses the concept o innovative packets to decidewhether a received packet contains new inormation or not.Simulation results show improvement in the total number o transmissions compared to ExOR. Opportunistic routing indynamic ad hoc networks (OPRAH) [] builds a threadedmultipath set between source and destination. It allowsintermediate nodes to have more paths back to the receiverand destination. However, duplicate packet reception is anassociated drawback o this protocol.

Resilient and opportunistic mesh routing (ROMER) []builds the mesh route or every packet. It assumes there isan existing technique to nd the minimum cost rom eachmesh router to the gateway. When a packet is sent rom amesh router to the gateway, the source mesh router needsto set a credit cost. Te overall cost to deliver the packet isthe minimum cost plus the credit cost to reach the gateway.Te probability that each intermediate router can orward apacket depends on the quality o the link to the parent router.Te best-link-quality intermediate node orwards the packetwith a probability o . Te other nodes send the packetswith the current rate o the considered link divided by thecurrent rate o the best link. However, the disadvantage o this protocol is that it has to rely on an existing schemeto nd the minimum cost rom each mesh router to the

gateway. Te directed transmission routing protocol (DRP)[] is a variant o ROMER. It adjusts the probability ata orwarder in a different way. I a node is sitting on theshortest path to the destination, it orwards each packet witha probability o . Otherwise, the probability is dependenton the extra distance to reach the destination. Te longerthe distance, the smaller the probability. Geographic randomorwarding (GeRaF) [] selects the orwarding nodes usinglocation inormation. Nodes closer to the destination havea higher priority. It adopts hop-by-hop orwarder selection.Te disadvantage o this protocol is the cost to acquire thelocation inormation. Coding-aware opportunistic routing(CORE) [] is an integration o conned interow network

coding and OR. It enables a node to orward a packet to thenext hop that leads to the most coding changes. Tis iterativeorwarder-by-orwarder mechanism signicantly improvescoding gain with a slightly increased protocol overhead.

Cooperative opportunistic routing in mobile ad hocnetworks (CORMAN) [] is a network layer solution to

opportunistic data transer in mobile ad hoc networks. Tisscheme broadens the applicability o ExOR to mobile mul-tihop wireless networks without relying on external sources.Moreover, it incurs smaller overhead than ExOR by includingshorter orwarder lists in data packets. o reduce the over-head in route calculation, they developed proactive sourcerouting [], which introduced a large-scale live updateto increase throughput and decrease delay rom orwarderlist adaptation. Tis provides robustness against link-quality

variation using small-scale retransmission.Simulation resultsshow that drastic improvement in packet delivery ratio andaverage delay is achieved, compared to ad hoc on-demanddistance vector.

Tis paper contributes to a new statistical analyticalmodel or studying traditional multihop and O. Te modelshows improvement in throughput and ewer transmissionsto successully deliver packets to their destination. Althoughmany analyses have been proposed, this work is uniquebecause we consider cases where the distance is known.Moreover, we develop an innovative generic Markov chainmodel o ourproposed method, which canbe applied to otherO scenarios. As ar as we know, this is the rst method thatstatistically ormulates and shows stability in our proposedO. We consider all possible probabilities or successuldata transmission rom source to destination. Using theproposed model, we compare opportunistic transmissionwith conventional multihop transmission, which determinesthe most reliable available multihop path. Evaluation resultsdemonstrate that the O outperorms the best traditionalmultihop transmission in successul delivery, number o transmissions, transmission power, number o intermediatenodes, and delay.

Te rest o the paper is structured as ollows. Section explains the system model. Section presents our proposedanalytical model,comparing opportunistic transmission withtraditional multihop transmission. Section demonstratesthe evaluation results based on the proposed statisticalanalytical model. Finally, Section provides the conclusionand discusses uture work.

2. System Model

Te system consists o sender node and receiver node .Te sender and receiver are at distance rom each other. Inthe literature, there are many geographical routing protocolsin which nodes know their location. Te distance can becalculated using the geometric coordinates and position o all the sensors []. Hence, we assume that is known. Tisis a mild and reasonable assumption. We keep the distancexed, and all the intermediate nodes are at an equal distancerom each other. Every intermediate node canrelay the packetto nodes within the communication range o it. Let signal


12/70


be transmitted rom given sender node to receiver in aRayleigh ading channel. Te probability density unction o received power can be written as

() = 1 −/ , ()

where is the average received power o the signal. Assumethat the mean power level alls off according to the powero the range /. is the product o transmitted signalpower, transmitter and receiver antenna gains, and systemloss. is a path loss exponent. is set to or simplicity.In case o static node distribution, is a constant.

For a given transmission rate, provided signal-to-noiseratio, the required received power at the receiver to decode apacket successully is given by

= log 1 + 0 , ()where the minimum required received power or successully decoding a packet is given by

= 0 2 −1 . ()Tereore, successul transmission probability is obtained

as ollows:

= ∫∞

−/,

= ∫∞

−/ = − −

/∞

= −/ = 1/ .

()

3. Multihop Transmission

In the multihop scenario, the probability o success can bewritten as

= + . ()

is the probability o successul transmission romsource to destination. is the probability o successultransmission rom sender to intermediate node, and isthe probability o successul transmission rom intermediatenode to receiver. Te total distance is xed, and intermediatenodes are equidistant. Tereore, the probability o success

will be

= −(/2)/ −(/2)/ . () or nodes is

= −((/(−1))/) . ()Figure describes the traditional multihop transmission,

where represents the intermediate nodes... Opportunistic Multihop Transmission. In O, the sendertransmits the packet with a list o possible orwarders andpriorities. Te destination hasthe highest priority, a node thatis nearest to the destination has the second highest priority,and so on. All the intermediate nodes can act as a relay and can orward the packet directly to the destination i itis in range; otherwise, the packet goes to the next highestpriority node. All the intermediate nodes will keep a copy o overheard packets.

I the highest priority node successully delivers packetsto the destination, then the other nodes will discard thepacket. Otherwise, the next highest priority node will try to deliver the packet to the destination. Te cumulativesuccess probability is the “success probability o the highestpriority node and success probability o the next highestpriority node, with the product o ailure probability o highest priority nodes with respect to this node.” In caseo ailure o all possible cases o the O, the last case is amultihop, and success probability is . Te receiver will sendacknowledgement afer successul delivery o the packet witha success probability o .

I the number o nodes is , then the equation remains thesame as in the direct case:

=

−

/ , i = 2−/+ −(/2)/ −(/2)/ 1− −/ , i = 3−1=2

−(((−)/(−1)) ((−1)/(−1))) ∏=2

− −((−+1)/(−1)) + −((/(−1)))−1(1−− ) −3∏=0

1− −(−−2)/(−1)2 i > 3

. ()

With nodes, the total number o hops is . Tetotal probability o success is the sum o the probability rom source to destination and the probability romintermediate relay node to destination , with the product

o ailure probability o direct transmission rom source todestination.

I the number o nodes is more than , this equationshows the overall probability o success or nodes, which is


13/70


Ps0 Ps1Ps2 Ps3 Ps4

s i1 i2 i3 i4 r

F : raditional multihop transmission.

s i1 i2 i3 i4 r

P0

P1

P2P3

P4

P5

P6

P7 P8 P9 P10 P11

F : Opportunistic transmission.

the recursive summation o all success probabilities withailure probabilities o all the higher priority nodes withrespect to that node, up to nodes.

Figure depicts the probability o success o all possibleroutes to the receiver.

.. Expected Number of Transmissions. Te expected num-ber o transmissions (EX) [] can be calculated as

EX = 1 . ()EX is inversely proportional to the probability o success.

.. Markov Chain Model for Opportunistic Transmission. is the state o a given packet at time . In the consideredproblem, state means the node where the current packet islocated.Te state transition diagram is shown in Figure . Tetransition rom current state to next state is

= +1 = | = . ()I , is the expected number o transitions until the

Markov chain, starting in state , returns to that state, then = 1, . ()

We are interested in ,. o calculate 1,1 transitionsrom state 1 till we return to state 1, we have

1,1 = 1, + 1,1, = 1,1 − 1. ()

1 2 3(i) 4 5(j) N

aij

aN1

a13

a12

F : State diagram or opportunistic transmission.

Te generic state transition matrix is

=

0 12 13 ⋅ ⋅ ⋅ 10 0 23 ⋅ ⋅ ⋅ 2... d ......

d...... d (−1)

1 0

,

= 123 ⋅ ⋅ ⋅ .

()

Te positive recurrent aperiodic states are called ergodic.For an irreducible ergodic Markov chain, lim→∞, existsand is independent o . Furthermore, let

= or =

or =

,

1 = 1 = , =

= 1≤


14/70


(i) = 3

3 = 1

13 + 2=2

∏1


15/70


Nexttransmission

Packet withorwarding list/priorities

Transmission

Nodespriority romlef toright

r

Yes No

Success

r − 1 r − 2 r − i r − j 3 2

F : System ow chart.

: Nodes versus distance or multihop transmission.

Nodes Distance/success probability

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

4. Results and Discussion

We set the ollowing parameters or our simulation or thisscenario. is set to , is ., and we increase the numbero relay nodes between source and receiver to see the impact

on the probability o success.Te impact o distance on the probability o success

or multihop is depicted in able . It can be seen that theprobability o success decreases with increasing distance ordifferent numbers o nodes. When the distance betweensender and receiver exceeds a certain threshold, it decreasesthe probability o success. Moreover, the signal amplitude isdecreased with increasing propagation distance. In case o ailure o transmission o multihop transmission, transmis-sion needs to ollow the same path again until the data issuccessully transmitted.

able shows the impact o distance on the probability o success or O. It is clearly seen that the probability o success

: Nodes versus distance or opportunistic transmission.

Nodes Distance

. . . .

. . . .

. . . . . . .

. . .

: E X or multihop transmission.

Nodes Distance/EX

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

: EX or opportunistic transmission.

Nodes Distance/EX

. . . .

. . . . .

. . . . .

. . . . .

. . . .

is reduced as the distance increases. I we compare the valueswith multihop transmission, the probability o success or Ois higher than multihop in all cases. Te probability o successdecreases with increasing distance, but the impact is very low or O. Te main reason is that i one o the possible pathsto the destination ails, there are other paths that can lead tosuccessul delivery o the data. Te number o paths to thedestination increases with an upsurge in the number o nodesin a network, which increases the probability o success.

Te impact o distance on the expected number o transmissions or multihop is shown in able . It can be seenthat EX increases in proportion to increasing distance.

Te expected number o transmissions or opportunistic

transmission is presented in able . It is clearly seen thatthe impact o distance results in more transmissions. Incomparison with multihop transmission, O requires ewertransmissions to successully deliver the data.

Figure shows the outcome o increasing transmissionpower on the probability o success. Te probability o successrises with increased power or multihop and opportunistictransmission. Opportunistic transmission outperorms mul-tihop transmission.

It is clearly seen rom Figure that the number o trans-missions decreases as we increase transmission power. Operorms better than traditional multihop transmission. Teoutcome shows that O is more efficient approach to deliver


16/70


4 6 8 10 12 14 16 18 200

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Multihop

Opportunistic

P r o

b a

b i l i t y o

f s u c c e s s

( P s

)

PTransmit (Pt)

F : Effect o transmission power on .

4 6 8 10 12 14 16 18 200

5

10

15

20

25

30

35

40

45

50

E x p e c t e d n u m

b e r o

f t r a n s m i s s i o n s

( E T X )

Multihop

Opportunistic

PTransmit (Pt)

F : Effect o transmission power on EX.

the data to the destination in ewer transmissions. Further, itreduces the energy consumption due to the less packet lossesand retransmissions. Tereore, O reduces the energy con-sumption and keeps the most important resource o sensorsor a long period o time or communications. Consequently,maximizing the lietime o the resource constrained cognitiveradio sensor networks. Hence, the overall perormance o thecognitive radio sensor networks is enhanced.

5. Conclusion

wo types o transmission have been studied in this paper,multihop and opportunistic. More specically, a xed-distance-based statistical model is proposed or multihopand O or cognitive radio sensor networks. Additionally,the unique generic Markov chain model is proposed toshow the stability o O. O shows improvement in reliably delivering the packet in ewer transmissions in contrast tomultihop transmission. Hence, O successully delivers thedata in an energy efficient way, increases the sensor’s lietime,and improves overall system perormance. It opens a new direction or multihop cognitive radio sensor networking-related research.

We will extend this statistical analysis or random-distance intermediate node scenarios. We will also work on across-layer protocol design by incorporating these statisticalanalyses.

Conflict of InterestsTe authors declare no conict o interests.

Acknowledgment

Tis work was supported by Basic Science ResearchProgram through the National Research Foundation o Korea (NRF) unded by the Ministry o Education (NRF-RDAA).

References

[] O. B. Akan, O. B. Karli, and O. Ergul, “Cognitive radio sensornetworks,” IEEE Network, vol. , no. , pp. –, .

[] I. F. Akyildiz, . Melodia, and K. R. Chowdhury, “A survey onwireless multimedia sensor networks,” Computer Networks, vol., no. , pp. –, .

[] S. Salim and S. Moh, “A robust and energy-efficient transportprotocol or cognitive radio sensor networks,” Sensors, vol. ,no. , pp. –, .

[] R. Rajmohan, “opology control and routing in ad hoc net-works: a survey,” ACM SIGACT News, vol. , no. , pp. –,.

[] I. F. Akyildiz, X. Wang, and W. Wang, “Wireless mesh networks:a survey,” Computer Networks, vol. , no. , pp. –, .

[] S. Lee, S. Banerjee, and B. Bhattacharjee, “Te case or amultihop wireless local area network,” in Proceedings of theIEEE International Conference on Computer Communications,pp. –, Hong Kong, March .

[] D. Aguayo, J. Bicket, S. Biswas, G. Judd, and R. Morris, “Link-level measurements rom an .b mesh network,” in Proceed-ings of the Conference on Applications, Technologies, Architec-tures, and Protocols for Computer Communications (SIGCOMM ’), pp. –, ACM, Portland, Ore, USA, August-September.

[] S. Biswas and R. Morris, “Opportunistic routing in multi-hopwireless networks,” ACM SIGCOMM Computer Communica-tion Review, vol. , no. , pp. –, .


17/70


[] S. Biswas and R. Morris, “ExOR: opportunistic multi-hoprouting or wireless networks,” in Proceedings of the Confer-ence on Applications, Technologies, Architectures, and Protocols for Computer Communications (SIGCOMM ’), pp. –,ACM, Philadelphia, Pa, USA, August .

[] J. Ai, A. A. Abouzeid, and Z. Ye, “Cross-layer optimal policiesor spatial diversity relaying in mobile ad hoc networks,” IEEE

Transactions on Wireless Communications, vol. , no. , pp.–, .

[] A. Bletsas, A. Khisti, D. P. Reed, and A. Lippman, “A simplecooperative diversity method based on network path selection,”IEEE Journal on Selected Areas in Communications, vol. , no., pp. –, .

[] H. Füßler, J. Widmer, M. Käsemann, M. Mauve, and H.Hartenstein, “Contention-Based orwarding or mobile ad hocnetworks,” Ad Hoc Networks, vol. , no. , pp. –, .

[] P. Larsson, “Selection diversity orwarding in a multihop packetradio network with ading channel and capture,” ACM SIGMO-BILE Mobile Computing and Communications Review,vol.,no., pp. –, .

[] R. C. Shah, A. Bonivento, D. Petrović, E. Lin, J. Van Greunen,and J. Rabaey, “Joint optimization o a protocol stack or sensornetworks,” in Proceedings of the IEEE Military CommunicationsConference (MILCOM ’), vol. , pp. –,IEEE, Monterey,Cali, USA, November .

[] B. Zhao and M. C. Valenti, “Practical relay networks: a gen-eralization o hybrid-ARQ,” IEEE Journal on Selected Areas inCommunications, vol. , no. , pp. –, .

[] M. Zorzi and R. R. Rao, “Geographic random orwarding(GeRaF) or ad hoc and sensor networks: energy and latency perormance,” IEEE Transactions on Mobile Computing , vol. ,no. , pp. –, .

[] K. Zeng, W. Lou, and M. Li, Multihop Wireless Networks:Opportunistic Routing , Wiley, .

[] M.-H. Lu, P. Steenkiste, and . Chen, “Video transmission overwireless multihop networks using opportunistic routing,” inProceedings of the Packet Video Workshop, pp. –, IEEE,Lausanne, Switzerland, November .

[] Z. Zhong, J. Wang, S. Nelakuditi, and G. Lu, “On selectiono candidates or opportunistic anypath orwarding,” ACM SIGMOBILE Mobile Computing and Communications Review, vol. , no. , pp. –, .

[] S. Chachulski, M. Jennings, S. Katti, and D. Katabi, “radingstructure or randomness in wireless opportunistic routing,”in Proceedings of the Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications(SIGCOMM ’), pp. –, ACM, Kyoto, Japan, August .

[] C. Westphal, “Opportunistic routing in dynamic ad hoc net-

works: the OPRAH protocol,” in Proceedings of the IEEEInternational Conference on Mobile Adhoc and Sensor Systems(MASS ’), pp. –, Vancouver, Canada, October .

[] Y. Yuan, H. Yang, S. H. Wong, S. Lu, and W. Arbaugh,“ROMER: resilient opportunistic mesh routing or wirelessmesh networks,” in Proceedings of the st IEEE Workshop onWireless Mesh Networks (WiMesh ’), Santa Clara, Cali, USA,September .

[] M. S. Nassr, J. Jun, S. J. Eidenbenz, A. A. Hansson, and A. M.Mielke, “Scalable and reliable sensor network routing: peror-mance study rom eld deployment,” in Proceedings of the thIEEE International Conference on Computer Communications(IEEE INFOCOM ’), pp. –, Anchorage, Alaska, USA,May .

[] Y. Yan, B. Zhang, J. Zheng, and J. Ma, “CORE: a coding-awareopportunistic routing mechanism or wireless mesh networks,”IEEE Wireless Communications, vol. , no. , pp. –, .

[] Z. Wang, Y. Chen, and C. Li, “CORMAN: a novel cooperativeopportunistic routing scheme in mobile ad hoc networks,” IEEE Journal on Selected Areas in Communications, vol. , no. , pp.–, .

[] Z. Wang, Y. Chen, and C. Li, “PSR: a lightweight proactivesource routing protocol or mobile ad hoc networks,” IEEETransactions on Vehicular Technology , vol. , no. , pp. –, .

[] G. Guido, A. Vitale, V. Astarita, F. Saccomanno, V. P. Gioŕe, andV. Gallelli, “Estimation o saety perormance measures romsmartphone sensors,” Procedia—Social and Behavioral Sciences, vol. , pp. –, .

[] D. S. J. De Couto, D. Aguayo, J. Bicket, and R. Morris, “Ahigh-throughput path metricor multi-hop wireless routing,” inProceedings of the th annual international conference on Mobilecomputing and networking (MobiCom ’), pp. –, SanDiego, Cali, USA, September .


18/70

Research Article A Cognitive-Radio-Based Method for Improving Availability inBody Sensor Networks

Olga León,1 Juan Hernández-Serrano,1 Carles Garrigues,2 and Helena Rifà-Pous2

Telematics Department, Technical University of Catalonia, Barcelona, SpainIT, Multimedia and Telecommunications Department, Open University of Catalonia, Barcelona, Spain

Correspondence should be addressed to Olga León; [email protected]

Received July ; Accepted September

Academic Editor: Gyanendra P. Joshi

Copyright © Olga León et al. Tis is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

One o the main threats to body sensor networks (BSNs) is Denial o Service attacks that disrupt communications used to transmitpatients’ health data. Te application o cognitive radio (CR) technology into BSNs can mitigate such a threatand improve network availability, by allowing network nodes to cooperatively agree on a new radio channel whenever the quality o the channel being inuse decreases. However, the cooperative spectrum sensing mechanisms used by CRs should also be protected to prevent an attackerrom predicting the new channel o operation. In this work, we present a lightweight and robust mechanism that appropriately secures the channel selection process while minimizing resources consumption, thus being suited or resource constrained devicessuch as body sensor nodes. Te proposed method has been analyzed in terms o energy consumption and transmission overheadand it has been shown that it outperorms existing cryptographic approaches.

1. Introduction

Sensor and wireless communication technologies are rapidly evolving and spreading to many elds, such as medicalservices. Body sensor networks (BSNs) [, ] are becomingmore popular and powerul every day and ongoing efforts,such as the IEEE .. standard optimized or low-powerBSN devices [], clearly reect the increasing importance and

potential o these types o networks.A typical BSN is composed o a number o sensors that

are placed at various locations on the body or in body,also known as implantable medical devices (IMDs). Asdepicted in Figure , these sensors orward sensed data toa more computationally powerul device or gateway (e.g.,a smartphone) that, in its turn, can transmit the gathereddata to a medical center. Tereore, the proessionals canconstantly monitor the patient’s state and take the properactions according to the observed data. Tus, the use o BSNs can considerably reduce the gap between a medicalemergency and the medical response while increasing theautonomy o patients, that is to say, their quality o lie.

Body sensors exhibit more constraints regarding size,power, battery availability, and transmission (i.e., the humanbody is a lossy medium) than those sensors that can beound in conventional wireless sensor networks (WSNs)and, thereore, they require specic solutions. Besides therecent IEEE .. standard, already supported by a ew commercial devices, several low-power wireless technologies[–] suitable or BSNs have emerged during the last years.

Tese technologies dene typical transmission rates rangingrom several kbps in AN+ to Mbps in WiFi with the lowestpower .b mode.

Lately, therehas been increasing concern in incorporatingsecurity and privacy mechanisms to medical systems in orderto preserve patients’ privacy and offer continuous monitoringo their health status. Besides, FDA (Food and Drug Admin-istration) made recently a call or manuacturers to addresscybersecurity issues relevant to medical devices or the entirelie cycle o the device []. Tus, it is expected that these actswill denitely encourage a number o works in this eld.

Generally speaking, the ollowing security servicesshould be provided in any medical system.




19/70


3G/4G

BSN

Gateway

Sensor

WiFi/Bluetooth

Internet Medical center

F : BSN model.

Condentiality . Data regarding patients’ state should be only accessible to authorized entities. In this context, this impliesthat only the BSN’s nodes should be able to interpret thesensed data.

Authentication. Te BSN’s nodes should be able to veriy thesource o any received data.

Integrity . Data should not be modied by an unauthorized

entity, or at least BSN’s nodes should be able to detect thatdata has been altered.

Availability . Data anddevice inormation should be accessibleupon request by authorized entities. Te human body is ahighly dynamic physical environment wherewireless channelproperties constantly change. Besides, these communicationscan be severely affected by intererences caused by electronicdevices in the proximity o the BSN.

Te rst three security goals can be easily achievedby means o classical cryptographic tools in conventionalnetworks. However, the limited capabilities o body sensorsmay prevent rom applyingthem to BSNs. Besides, traditional

cryptographic tools cannot prevent disruption o the network services due to intererences, no matter whether they areintentional, or example, a Denial o Service (DoS) attack, ornot. Given the relevance o the data sent by body sensors,there is clearly a need or mechanisms to maximize theavailability o such networks.

Te integration o CR technology [–] into BSNs, lead-ingto the concept o cognitivebody sensornetworks(CBSNs)[], can signicantly improve availability by allowing thenodes to select the best channel at any moment and avoidthe harmul effect o intererences. CRs exchange sensed dataabout channel availability and jointly agree to switch to a new channel when the channel being in use becomes unavailable.

Note that i an attacker manages to eavesdrop channelavailability data, it can take advantage o it to perorm a new attack on the new channel o operation, thus preventing thenetwork rom using an available channel and leading thenetwork to a DoS []. Channel switching, i unpredictable,renders DoS attacks more difficult since the attackermust jamevery possible transmission channel. raditional encryptionand authentication o exchanged data may help to hide chan-nel switching decisions rom external attackers but entail an

additionalcost thatcannot be assumed by heavily constraineddevices such as IMDs.

In this paper we present a protocol to protect the processo channel selection in CR-based BSNs. Te main goal isto maximize the availability o the network, thus ensuringthat patients’ data such as blood pressure, heart rate, andtemperature will successully be delivered to a gateway (nonstop monitoring o patients). Te protocol makes use o lightweight encryption and authentication primitives speci-ically suited or constrained devices such as body sensors.

Te main contributions o this paper can be summarizedas ollows.

(i) We apply CR technology into BSNs in order to

maximize the availability o services in such networks.Because CRs are able to sense the medium and selectthe best transmission channel at any moment, theeffect o intererences or DoS attacks can be miti-gated. In a conventional network, such phenomenawould interrupt communications within the BSNs.In a CBSN, the nodes can switch to a new channelwhenever the channel in use becomes unavailable.

(ii) We propose a method, suited to constrained devicesas body sensors, to secure the exchange o channelavailability inormation and prevent an attacker romeavesdropping such data, thus diminishing the prob-ability o a successul DoS attack.


20/70


(iii) We provide a security analysis o the proposedmethod and derive the time period during which thecryptographic material remains secure.

(iv) Te proposed method is compared to otherapproaches based on traditional cryptographic prim-itives in terms o energy consumption and CPU

usage.

Te rest o this document is structured as ollows. InSection we review the state o the art on security in BSNs.Section describes the BSN model considered in this work and its potential threats. A lightweight method to securethe process o channel selection in a BSN is presented inSection . Sections and present a security analysis o theproposed method anda comparison with existing approachesin terms o resources consumption. Finally, in Section weprovide the conclusions o this work.

2. Related Work

o date, research on security in BSNs has mainly ocused onprotecting data stored at the network nodes rom unautho-rized access and providing authentication and condentiality to the communications among the BSN devices. In theollowing, we provide an overview o the proposals that canbe ound in the literature.

Many proposed authentication methods are based onbiometrics, that is, relying on measurements o physiological

values (PVs) [], such as heart rate, blood pressure, ortemperature, in order to establish trust and generate key material. Te main idea is ensuring access to sensors only to those devices in physical contact with the patient. Teadvantage o these methods is that the key source is hard oran attacker to predict without physical access to the patientand also ensures orward-security, because PVs change overtime. Te main challenge, however, is how to achieve success-ul authentication among authorized devices when the PVmeasured by each one is not exactly the same, either due tomeasurement errors or due to the act that different devicesmeasure a given PV at different time instants.

Authentication by means o distance-bounding protocolswas proposed in several works [, ]. Tis techniqueprovides a very weak mutual authentication between twodevices based on measuring the transmission time betweenthem. Te rationale behind these protocols is that a legitimatedevicemust be closerthan a given distance.As a consequence,

they are vulnerable to injection attacks as long as the attackeris close enough to the patient bearing the sensors, orexample, by means o a hug.

In [], the authors presented a protocol based onidentity-based encryption (IBE). IBE systems are public key cryptosystems that allow any device to generate a publickey rom a known identity value such as the sensor IDand require the existence o a trusted third party called theprivate key generator (PKG) to generate the correspondingprivate key. o reduce the burden o key generation andencryption/decryption introduced by traditional public key cryptography, the authors proposed to use elliptic curvecryptography (ECC), which provides public key primitives

suitable or constrained devices as sensors in BSNs. Despiteit, it is still more expensive in terms o resource consumptionthan approaches based on symmetric cryptography.

In order to preserve user’s privacy, a number o worksproposed the use o symmetric encryption based on the AES(Advanced Encryption Standard) algorithm [–]. Many,

such as the one in [], proposed to use AES with CCMmode o operation, that is to say, AES counter (CR) modeor data encryption and AES cipher-block-chaining messageauthentication code (CBC-MAC)or message authentication.Te main advantage o this mode is that the same key can be used or authentication and or encryption withoutcompromising security and there is no need or rekeyingas long as the number o devices is xed. As a drawback,the added cost o encryption/decryption and, especially, thecosts due to the transmission overheads cannot be neglectedin BSNs, where every step orward in resources’ saving iso paramount importance. In this line, the authors in []presented an in-network mechanism that mimics the AESalgorithm and greatly reduces the costs o decryption whilethey claim achieving the same level o security.

All the above-mentioned proposals approach the prob-lem o protecting patient’s data rom unauthorized access,modication, or orgery but cannot effectively deal with DoSattacks. Such a protection can be achieved by making use o CR devices that collaboratively switch to another requency band [, , ] i the signal-to-noise ratio o the current oneis below the required value. Furthermore, it is also necessary to protect the exchanged sensing data in order to prevent anattacker rom eavesdropping data and get the next channelto be used in the network. Note that this inormation may allow an attacker to rapidly perorm a DoS attack in the new channel.

In this work, we present a lightweight and secure methodthat makes use o CR technology or improving the availabil-ity o the system, that is, ensuring that the communicationbetween the body sensor nodes will be available even underthe presence o unintentional or intentional intererences.Te application o CR technology into body sensor networkswas already proposed in previous works [, ]. However, tothe best o our knowledge, none o them addressed security topics.

In [, ], several methods or securing spectrum sens-ing mechanisms were discussed, but they are not suited orheavily constrained devices such as body sensors.

In [], the authors aimed to improve the availability o

a BSN by means o a cross-layer multihop protocol that dealtwith routing o data. Tis scheme, however, can be appliedonly to multihop BSNs where the path between two givennodes is established according to the connectivity among thenodes. In this approach, nodes make use o several paths butone single channel and thus are more vulnerable to attackssuch as jamming than CR-based networks.

3. Network Model and Threats

In this work, we have considered a BSN composed o aset o sensor nodes where all o them can act as sinks,collecting/storing data rom other sensors and potentially


21/70


Gateway

Sensor

Sink

F : Communication between sensors and the gateway.

transmitting these data to an external gateway i required (see

Figure ). Although this approach introduces some overheaddue to the act that data must be shared among all sensors,it improves network availability and robustness against dataloss and airly distributes energy consumption among allsensors. Also,it makes the process o gathering by the gateway easy, which can connect to any o the BSN nodes to get all theinormation.

As previously mentioned, we also assume that sensorshave cognitive capabilities; that is, they orm a CBSN andare able to identiy ree spectrum bands and adapt theirtransmission parameters accordingly. Spectrum sensing canbe perormed by each node on an individual basis or cooper-atively. As the latter increases the probability o detection due

to space diversity [], we have adopted such an approach inthis work.In cooperative spectrum sensing, each sensor senses

the medium and exchanges its observations with the othermembers o the network in order to agree on a given channelor data transmission/reception. However, these control dataare exposed to many attacks [], such as packet injection,eavesdropping, or Denial o Service (DoS). Next, we describethe attacker model and the specic attacks that can beexecuted against CBSNs.

.. Attacker Model. In this work we ocus on outsiders, thatis, external attackers that do not share any cryptographic

content with the gateway or the victim’s sensor nodes. I the attacker nodes are part o the CBSN, they will haveaccess to the keying material and thereore will be able tosuccessully eavesdrop and inject data. In any case, the designo a mechanism to counteract this threat is out o the scopeo this work.

In the context o CBSNs, we can classiy adversariesaccording to the ollowing criteria:

(i) Activeor passive: a passive attackercan only eavesdropdata, thus being able to access patient’s data and

violating his/her privacy. In its turn, an active attackeraims at injecting or modiying data in order to sendake reports on the state o the patient.

(ii) Type of attack includes the ollowing:

(a) eavesdropping: unauthorized access to storeddata or to transmitted data among the CBSNdevices, thus violating the privacy o the patient,

(b) modication/injection: an attacker that may alter the content o a packet transmitted by asensor or impersonate a sensor by orging apacket; these attacks can be executed due to lack o authentication and violate the integrity o theCBSN communications,

(c) packet replay: an attacker that may capture apacket that was previously sent by a sensor o the network. Regardless o the act that theCBSN is using authentication mechanisms ornot, the packet will be accepted by the networksi antireplay mechanisms are not provided,

(d) jamming: the adversary that disrupts the CBSNcommunications by generating interering sig-nals.

(iii) Intentional or unintentional : the adversary can bean external entity willing to cause damage to thecommunications among sensors and the gateway orcan be an entity that unintentionally is causing inter-erences to those communications. As an example,the patient o interest could be near another patientwith wearable sensors, which could injectake reportsi data is not properly authenticated. Examples o unintentional attacks could take place in a situationwhere two patients bearing body sensors are huggingand unconsciously exchange data. Or the patientcould be near a relative who is visiting him/her at the

hospital and carries any electronic device that causesintererences to the CBSN.

It is important to remark that, in a CBSN where sensornodes establish communications using different channelsover time, these attacks can be extended to the control dataexchange among the devices o the CBSN. As an example, anattacker may orge a report regarding the availability o thechannels, thus leading the CBSN to select a channel that issuffering rom high intererences or that is currently beingused by another service. Note that this attack can lead to aDoS and the ailure o the system in monitoring the patient’sstatus. In itsturn, eavesdropping o the control channel allowsan attacker to have knowledge o the channels to be used

by the CBSN. Te attacker could take advantage o thissituation in order to easily disrupt the communications inthe network by perorming a new DoS attack every time theCBSN switches to a new channel.

Te implementation o security mechanisms in a CBSN[] to counteract these attacks is specially challenging dueto the limited capabilities o CBSN’s nodes. In the ollowingsection, we describe a simple method to secure the processo channel selection in CBSNs. Te proposed mechanismis suited or networks with extremely constrained-resourcesdevices, since it makes use o lightweight cryptographicunctions and minimizes the added transmission/receptionoverhead.


22/70


4. Securing Sensing Data andChannel Selection in CBSNs

In the ollowing we present a mechanism or securing theexchange o sensing data and the channel selection processin CBSNs. Section . outlines the assumptions considered in

this work regarding the network model and, in Section ., wedescribe the protocol operation. For ease o understanding,we present the terminology used along this section as ollows:

CR : medium-term session counter ( bits),

CR : short-term session counter ( bits),

: data sensed by node during period ( bits),

ID: link-layer identier o node ( bits),

: keystream to encrypt and authenticate data ornode during period ( bits),

KM: keying master,

: length o the data sensed by a given node during a

given period,

: length o the hash output and all the secrets,

: number o nodes in the network,

: number o keystreams obtained rom a ( =

/), dening the number o sensing periods beoreupdating ,

: length o the keystreams , which must be adivisor o ,

: long-term globally shared secret (-bits),

: medium-term globally shared secret (-bits),

,: long-term secret shared between the KM andnode ; it is used to update in case it is compro-mised,

: short-term shared secret with node .

.. Assumptions. Although the proposed protocol isdesigned to be implemented in heavily constrained devices,we work under the assumption that such devices have at leastthe ollowing capabilities:

(i) Compute a hash unction with an output length o bits.

(ii) emporally store in itsrandom accessmemory at least

⋅ ( + 3) bits, with the number o nodes in thenetwork. As we detail later in Section ., each nodemust keep a short-term shared secret or each o theN nodes in the network (including itsel) and threemore long-term and medium-term secrets, each onewith length o bits.

(iii) Sensor nodes use a synchronization protocol that willbe used to share a global short-term session counterand a medium-term session counter among all nodes(see Section .). Given the low transmission rate o sensor networks, existing synchronization schemes[] provide enough precision or this purpose. Weassume that the chosen protocol provides recovery

methods upon loss o synchronization. How syn-chronization is achieved will strongly depend on thechosen protocol, but i the latter requires a masternode or providing synchronization, the gateway o the BSN could play this role.

o the best o our knowledge, the ormer requirement canbe assumed even in very constrained devices. As shown in[], there are several lightweight hash unctions that canbe integrated into a sensor mote. Te latter may not beharder to achieve. As detailed in Section ., during every sensing period each node stores one secret per member o the network, a globally shared secret, and two counters, allo them with the same length as the hash output. I weconsider a typical hash unction with an output o or bits and a network with tens to hundreds o sensors, theRAMrequirements or sensor nodes are just bounded to a ew tenso kilobytes.

.. Protocol Operation. Beore deploying the CBSN, every sensor node must be preloaded by a keying master (KM) withthe ollowing data:

() Te set o channels that the sensor will have to sensein the cooperative sensing process.

() A long-term and globally shared secret o bits(the hash output length).

() A long-term secret , shared between the KM andnode that will be used to update the globally sharedsecret in case it is compromised.

Te KM is an external device, which is not a member o theBSN. ypically, this role is playedby thedeviceresponsibleor gathering data rom the sensors or gateway (e.g., a smartphone or a tablet).

Upon deployment o the network, every node derives amedium-term globally shared secret by hashing the XOR o the long-term secret and a counter. Te generationprocess o is clearly depicted in Figure . Tis process isperiodically repeated with an updated value o the medium-term counter in order to protect the secret against a potentialattacker. Details about how ofen this process should becarried out and the attacker capabilities are provided inSection .

As shown in Figure , each node generates a set o random sequences o bits: one or the node itsel and one

or each other node in the network. Tese random sequences,, with the node identier, are obtained by hashing theXOR o the link-layer identier o the node ID, the medium-term shared secret , and a short-term session counterCR .

Tereore, ourproposal makes useo three types o sharedsecrets:

(i) A short-term per-node shared secret (one per eachnode in the network) used orencryption, decryption,and authentication o data.

(ii) A medium-term globally shared secret thatis usedto derive the short-term per-node shared secrets .


23/70


Long-term shared secretSL

Medium-term session counter

Medium-term shared secret

CTR M

Hash

SM

m bits

m bits

m bits

m bits

F : Generation o the medium-term globally shared secret .

Short-term session counterMedium-term shared secret

node uShort-term shared secret with

Hash

SM

Su

Identier of node uIDu

CTR S

m bits

m bits

m bits

m bits

m bits

F : Generation o the short-term shared secret or node .

(iii) A long-term globally shared secret that is used toderive a new when the current one is about toexpire.

Asclearly denoted in Figure , each sequence is dividedinto ragments o bits, which we will be denoted as ,each one being used as keystream to encrypt and authenticatedata or node in period . As per this behavior, a new short-term shared secret must be derived every sensing period.

When a node perorms spectrum sensing, it generates abinary sequence o bits that stores the availability o thedifferent channels. Te length o such sequence will dependon the number o bits used to code the state o each channeland the number o channels. As an example, the simplest way would be to use just a single bit or coding each channel, with

value “” i the channel is occupied and “” otherwise. I moreprecise inormation about the quality o channels is needed

(i.e., high, medium, low, and very low quality), more bits canbe used to code the channel state.

During a sensing period , each node must send toits neighbors its own sensing inormation but also it mustprocess the inormation received rom its neighbors to reacha joint decision.

In order to send its own sensing inormation, node willmake use o the corresponding keystream : the rst bitso the keystream will be used to encrypt channel inormation by means o a XOR addition; the remaining − bits arelef unchanged and will be used to provide message authen-tication, as illustrated in Figure . Te resulting sequence will be sent to all the other nodes.

o veriy the authenticity and decrypt the content o thepackets that have been sent by a given neighbor , a nodewill XOR the sequence o the received packet with the

keystream , as depicted in Figure . I the last − bitso the resulting sequence are not all s, the authenticationails and the entire packet is discarded. Otherwise, channelinormation can be recovered rom the rst bits resultingrom the XOR addition.

Te above described process is applied or each neighbor-ing node . Ten, the channel reported by a larger number o neighbors will be selected or the operation o the network.Note that because more than one channel may be reported by the same amount o nodes, a tie-break mechanism is neededto guarantee that the process leads to equal results in allnodes. One simple approach that could be used is to selectthe channel with the highest identier. However, this would

lead to a lower usage o channels with lower identiers andthereore to providing the attacker with valuable inormationabout channel usage in the CBSN. As a consequence, wepropose to use a tie-break method that relies on the ormato .

Recall that a undamental characteristic o this protocolis that there is no central entity that is known and trusted by all sensors. Tis makes the protocol suitable or unattendedscenarios, and it also makes it more efficient in terms o datatransmitted through the network, because no inormation issent regarding which channels have to be sensed or whichchannel is nally selected. Instead, sensors are deployedwith all the inormation needed to perorm the sensing in


24/70


Short-term shared secret with node u

Keystream pKeystream 2Keystream 1

Su

Encrypted Auth.datasensing data

Sensing data

Sensing data

Sensing period 1

Sensing period p

Encrypted Auth.datasensing data

...

Cup

Cu1

Dup

Du1

Ku1 Ku2 K

up

m bits

r bitsr bitsr bits

l bitsl bits

l bitsl bits

l bits

l bits

l bitsl bits r-lbitsr-lbits

r-lbits

r-lbits

F : Encrypting and authenticating sensing data.

Short-term shared secret with node u

Keystream pKeystream 2Keystream 1

Su

Sensing data

Sensing period 1 True

Accept sensing data

Reject sensing data

...

Cu1

Du1

Ku1 Ku2 K

up

If ≠ 0s

r bitsr bitsr bits

r bits l bits

m bits

r-lbits

F : Decrypting sensing data.

a distributed way and make a joint decision autonomously.Tus, there is no need or additional mechanisms to be used

when a new node joins the network. In this case, the new node needs to synchronize with the rest o the members to getthe proper value o the session counters by making use o thecorresponding protocol. However, when a node is expelledrom the network because it has been compromised, new cryptographic material must be generated and distributedamong the remaining nodes. Te KM is responsible ortriggering this process and communicates with each sensornode to update the shared long-term secret . Note thatbecause the KM shares a different secret , with each node ,it can securely distribute the new value o . Upon receptiono , each node should perorm again the initializationprocess described at the beginning o this section.

5. Security Analysis

Te security o the proposed method relies on the sharedsecrets used to derive the keys and perorm encryption andauthentication o channel availability data. As long as thesesecrets are not compromised, data condentiality can beensured; that is, an attacker might not be able to get the listo channels to be used in the CBSN. Besides, the methodmust prevent an attacker rom injecting ake data into thesystem. Tese issues are discussed as ollows. In Section .,we analyze how ofen the shared secrets should be updated inorder to guarantee a proper protection against cryptanalysis;next, in Section ., we evaluate the packet authenticationmethod used in our proposal in terms o probability o bypassing the authentication check.


25/70


.. Shared Secrets Lifetime. As previously mentioned inSection ., or this analysis, we are assuming that attackscome rom external entities and thereore attackers are notable to obtain the cryptographic material that is stored in thebody sensors. In the context o this proposal, the lietime o each o the shared secrets is the interval in which these secrets

are considered computationally sae against cryptanalysis,that is, their cryptoperiod.Te cryptoperiod straightly depends on the chosen cryp-

tographic protocols, the length o the secrets themselves, andthe amount o times they are used. Te more a given secret isused, the shorter its cryptoperiod is, as an attacker gets moreinormation about this secret and thereore the probability o a successul cryptanalysis increases. In act, cryptoperiod isdened more in terms o the number o times a given secretor key is reused (the amount o ciphertext exposed to anattacker or a given secret/key) than as a given time period,which strongly depends on the transmissionrate o the sensornodes.

Recall the three types o shared secrets used in theproposed method:

(i) Short-term per-node shared secrets: one secret persource node that is used to lightweight encrypt,decrypt, and authenticate the channels’ sensed data.

(ii) Medium-term globally shared secrets: globally sharedsecret that is used to derive the short-term per-node shared secrets .

(iii) Long-term globally shared secret : this being the ini-tially preloaded secret that is used to derive a new medium-term globally shared secret when thecurrent one is about to expire.

As clearly denoted in Figure , our approach operates, insome manner, as an additive stream cipher. It is well knownthat stream ciphers are considered to be secure as long as thekey is never reused and, thus, our cipher will be secure i agiven value is not repeated. As a result, must be updatedevery = / sensing period, with the renewal period, the length o the shared secret , and the amount o transmitted bits (sensing data) in a sensing period that areencrypted with .

Recall that, in our proposal, the per-node key used orencryption is generated by means o a hash unction. Asecond requirement is that this unction must be crypto-graphically secure. Note that i the hash unction does not

accomplish it, an attacker mightbe able to reverse it, thatis, toget the input o the hash unction given an output, meaningthat, in our proposal, an attacker would be able to recoverthe value o the medium-term globally shared secret (seeFigure ).

A cryptographically secure hash unction with an outputo bits can offer a security level o 2 operations against

preimage attacks and 2/2 against collision attacks. Generally speaking, a minimum output o bits is required in orderto provide a high level o security or most applicationsbut shorter lengths are accepted i the number o generatedmessages in a given period is limited, as it is the case o low-rate networks. In Section , we propose several lightweight

hash candidates with an output o bits; that is to say,we can assume that it is computationally uneasible or anattacker to invert the hash unction and thus to predict the

value o as long as it is updated beore exceeding its

cryptoperiod, which has an upper bound o 2/2 = 264 uses.Te long-term globally shared secret is only used to

update the current medium-term globally shared secret .Because is not updated very ofen, it is very unlikely thatan attacker manages to obtain several values o to reversethe hash unction and recover . As a result, we can assumethat the cryptoperiod is long enough and there is no needto update the secret during the nodes’ lietime.

.. Authentication. A cryptogram o sensing data con-tains an authentication eld o bits that is checked uponreception (see Figure ). Consequently, an attacker has a

chance o in 216 o guessing the next authentication eld,which allows it to orge a valid authentication eld and injectake data. Note that this attack can lead the CBSN to wrong

decisions about the availability o the spectrum.I the attacker repeatedly attempts to send valid cipher-

texts, it may succeed afer 215 attempts, in average. Becausethe attacker does not know , the authentication eldappears to it as a random stream and thereore it must select aake authentication eld at random. Besides, the attacker can-not determine whether a given ciphertext has been acceptedor rejected because the receiver does not acknowledge thereception o such packets to the emitter. Otherwise, theattacker could take advantage o this inormation in order toguess a valid authentication eld in a aster way.

In conventional networks, 215 packets may seem anextremely low number but it may provide an adequate level

o security in CBSNs. In these networks, the attacker can only send ake packets during the sensing periods, which is in theorder o a ew milliseconds in most cognitive scenarios [].Moreover, as previously stated in Section , transmission ratesin BSNs are considerably low, with values usually rangingrom tens to a ew hundred o kilobits per second.

As an example, let us consider a Mbps link, a sensingperiod o ms, and a packet size o bytes (which is clearly bigger than the typical packet size in sensor networks). Giventhese parameters, an attacker would only be able to send packets at most in every sensing period. Tat is to say, theattacker would need an average o . sensing periods tosend a ake packet and pass the authentication check.

6. Cost Evaluation and Comparison with Other Approaches

In this section, we evaluate the cost o our proposal interms o energy consumption due to transmission overheadand computational cost and compare its perormance withthe most common approach adopted in sensor networks[], which is providing authentication and/or encryption o the channel sensing data by means o using standard block ciphers. As is well known, block ciphers have as input themessage to be encrypted or authenticated, which is dividedinto several blocks o x length and a key. Both the block


26/70


length and the key length depend on the algorithm beingused. Regardless o the algorithm, block ciphers can be usedin several modes o operation depending on the service tobe provided, that is, encryption only, authentication only, orencryption/authentication. Generally, the ollowing modes o operation are applied.

Authentication. CBC-MAC is a block cipher mode or gen-erating message authentication codes. Te message to beauthenticated is divided into several blocks o equal size, andeach block is encrypted so that the value o a given block depends on the encryption o the previous block. Te naloutput o the cipher, that is, the message authentication codeor CBC-MAC, is the result o encrypting the last block o themessage. When the input o the cipher is shorter than theblock size (as it is usually the case in sensor networks), theCBC-MAC can be obtained by directly encrypting a singleblock, padded until the block size o the cipher is reached.

Encryption. CR mode o operation turns a block cipher into

a stream cipher, meaning that the resulting ciphertext has thesame size as the input or plain text. Tus, it does not orce theoutput length to be a multiple o the block size, as it is the caseo other modes such as CBC-MAC. Tis property makes thismode o operation suitable or encryption in sensor networkswhere devices usually exchange short-length messages.

Authentication+Encryption. CCM (CR + CBC-MAC) is acommon choice or providing both encryption (CR) andauthentication (CBC-MAC) []. A minor variation o CCM,called CCM∗, is used in the ZigBee standard [].

In this work, we assume that Advanced EncryptionStandard (AES) in CR mode is used or encryption and AES

CBC-MAC or authentication, as it is the current standardor symmetric cryptography, even in sensor nodes [].Currently, there are efficient hardware implementations o AES that are highly affordable.

Regarding hardware platorms, the vast majority o previous works on BSNs have used the -bit exas Instruments’(I) MSP and CC amilies o microcontrollers [].Built around a -bit CPU, ultra-low-power MSP micro-controllers are designed or low cost and, specically, low-power consumption embedded applications. As an example,I’s CC amily [] enable robust Bluetooth low energy (BLE) network nodes to be built with low total bill-o-material (BOM) costs. BLE operates in the same spectrum

range (. GHz–. GHz ISM band) as classic Blue-tooth technology but uses a different set o channels; insteado MHz channels, BLE offers MHz channels, oradvertising purposes and or data exchange.

.. Transmission/Reception Overhead. In this section weanalyze the overhead introduced by our proposal in termso transmission/reception o channel availability data andcompare it with the overhead exhibited when conventionalapproaches are used or data encryption/authentication, asexplained above. We assume that all sensors are capable o sensing a given set o channels and report inormation abouttheir state.

With our proposal, the minimum number o transmittedbits willdepend on thenumber o channels that a given sensoris reporting, the number o bits used to code the state o each channel, and the length o the authentication code. Asexplained in Section ., a length o bits is enough to securemost applications in WSNs and, thus, we have assumed this

value or the authentication eld. Tis leads to a total amounto Bits = + 16 o transmitted bits, where represents thetotal number o bits used to code all possible channels, andBits = ( − 1)Bits received bits, representing the numbero bits received by a given sensor rom its neighbors.

Aiming to provide a air comparison, we choose the samekey length or block ciphers and or our proposal, that is tosay, a -bit key. Te transmitted bits overhead added by AES CBC-MAC authentication is bits (or a semantically secure implementation also an IV or nonce must be sharedbetween emitter and receiver, so that the overhead can behigher). Regarding encryption, the number o transmittedbits isequal to the number o bits usedtocodethestateothechannels, but it also requires the use o a nonce, with a lengthequal to hal o the key length, that is, bits per message.

During every sensing period, every node must transmita packet with sensing inormation but also must process thepackets received rom its neighbors. able and Figure ,respectively, show the transmission and reception overheaddue to the secure sharing o sensing inormation using bothstandard block ciphers and our proposal. Te values areprovided as a unction o the number o bits used to codethe state o the channels and the number o nodes , rangingrom to . Given that the considered scenario is a body sensor network, this is more than a reasonable value, sincea patient wearing more than sensors may be an unlikely situation.

Te reader may notice that the overhead introducedby this mechanism increase

Date post:	08-Jul-2018
Category:	Documents
Upload:	radhakodirekka8732
View:	216 times
Download:	0 times

Applications of Cognitive Radio Networks Recent Advances and Future Diections

Documents