Mobile gateway for ubiquitous health care systemusing ZigBee and Bluetooth

Teemu H. LaineDep of Information and Computer Engineering

Ajou UniversitySuwon, South Korea

Email: teemu@ubilife.net

Chaewoo LeeDept of Electrical Engineering

Ajou UniversitySuwon, South Korea

Email: cwlee@ajou.ac.kr

Haejung SukDept of Digital Media

Ajou UniversitySuwon, South Korea

Email: dbdip@ajou.ac.kr

Abstract—A ubiquitous health care system takes the advantageof portability and small size of wireless sensor nodes to provideremote health care services and real-time health monitoring.A gateway is needed to mediate communication between alocal sensor network and remote data consumers. In currentimplementations of ubiquitous health care systems ZigBee-basedsensors are often used to gather vital sign data such as ECGand heart rate. Transferring large quantities of vital signs from aZigBee-based sensor sink node to a gateway requires a bandwidththat ZigBee alone is not able to provide. In this paper, wepresent a technical design of a Bluetooth-based mobile gatewaythat bridges the connection between a sensor network and theInternet. Our system enables ubiquitous health care experiencewhile providing a platform for additional services such as alarms,notifications and analysis of medical data. Controlling a sensornetwork from the mobile gateway is also possible. The flexibledesign of the system does not restrict its usage only to health careservices – the gateway can be configured to work with any kindof sensor network having a sink node with Bluetooth capability.

I. INTRODUCTION

Ubiquitous systems are currently under active research andtheir popularity has been facilitated by increasing availabilityand affordability of powerful mobile devices and context-aware technology. Ubiquitous computing systems have beenemployed for example in education [1], [2], military [3], [4],transportation and supply chains [5], [6], and tourism [7], [8].Common characteristics for most of the ubiquitous systemsis the high user/object mobility and utilization of sensor andwireless communication technologies which are unobtrusiveto the users. A major advantage of ubiquitous systems isthat the users are able to access contextual and personalizedinformation regardless of location and time.

Wireless sensor networks achieve context detection in ubiq-uitous systems through the use of small sensor devices oflow cost and low power. Such networks are comprised ofseveral sensor nodes, each having a microprocessor, sensorchip(s), possible actuator(s) and wired/wireless transceiver.The sensor nodes provide intelligence information and usefulservices through negotiation with their neighbor nodes overa short range transmission protocol such as ZigBee. Wirelesssensor networks can be utilized widely in application fieldssuch as medical care, military tactics, home networking andenvironment monitoring. The concept of Body Area Network

(BAN) has emerged to support human body monitoring inhealth care and wellness services. In a BAN, wireless sensorsattached to a body collect different types of vital signs (heartrate, blood pressure, body temperature, etc). These sensorsare small in size, light-weight and wearable, thus enablingubiquitous health monitoring. BANs are in key role whenubiquitous health care systems are built.

Today mobile devices present highly potential client plat-forms for ubiquitous systems. Large screen sizes and gesture-based input methods on these devices enable rich presentationand interaction. At the same time processing and storagecapacity have grown to run smoothly 3D graphics and highquality videos. High-speed Internet access (4G, 3G, WiFi),Bluetooth, GPS and embedded sensors are available in mosthigh-end mobile devices and they are getting increasinglypopular in the low-end category as well. With these featuresat hand, we seek to create a general purpose mobile gatewayfor sensor-based ubiquitous systems. We demonstrate thegateway’s operability with a BAN-based ubiquitous health caresystem (U-Healthcare). In the proposed U-Healthcare systemour mobile gateway mediates communication between a BANand a remote server using ZigBee, Bluetooth and WLAN/3Gcommunication channels. Flexible architecture makes it pos-sible to add new sensor types to the system so it can be usedfor other purposes than ubiquitous health care as well.

This paper is organized as follows. We start by reviewingexisting research related to ubiquitous systems and previousattempts to combine Bluetooth and wireless sensors. Then wedescribe the basic concept of the proposed mobile gatewayfollowed by technical design where we explain the technicalarchitecture and communication aspects of the system. Finally,we discuss the findings, conclude the paper and present futureresearch challenges.

II. RELATED WORK AND DESIGN REQUIREMENTS

In this paper we demonstrate our mobile gateway archi-tecture with a ubiquitous health care system. Such systemscombine ubiquitous computing technologies with medical ser-vices through collecting, analyzing, delivering and managingvital signs regardless of time and space. Ubiquitous healthcare systems are particularly useful for providing remote clinicservices for patients located far from medical facilities. Based

2014 Eighth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing

978-1-4799-4331-9/14 $31.00 © 2014 IEEE

DOI 10.1109/IMIS.2014.17

139

on our analysis of existing health care systems, a typical archi-tecture is based on a layered model (Figure 1) comprising fouror more layers. The first layer is takes care of context sensingusing a Body Area Network (BAN) which comprises a numberof small sensor devices capable of measuring properties of ahuman body such as heart rate or oxygen saturation. Datacollected by the first layer is transported to the second layerover a short-range communication protocol such as ZigBeeor Bluetooth. The second layer is represented by devices oflimited processing power such as smartphones, laptops or set-top boxes. These devices operate as gateways compiling andpreprocessing (e.g. noise reduction filtering) the data beforesending them over the internet to the third layer. Additionally,the second layer may cache the data locally for visualization.The third layer acts as a central repository for data collectedfrom multiple gateways. It may provide advanced tools for dataretrieval and analysis as well as interfaces for the presentationlayer. The fourth layer, the presentation layer, offers multiplemeans of accessing, visualizing and manipulating data byhealth care professionals and other stakeholders. Although wedecribe this generic architecture for ubiquitous health caresystems, the same layered model can be applied to many otherubiquitous system architectures.

Fig. 1. Generic layered architecture for ubiquitous health care systems

Based on the aforementioned layers we can draw designrequirements for ubiquitous health care systems. Firstly, thesystem should provide reliable medical services that comple-ment services provided at medical facilities. Secondly, patientsshould receive these medical services conveniently with aminimal impact on their work and common life activities.Thirdly, a high-speed wireless network infrastructure shouldbe present to support low-latency data transmission betweenparties involved (i.e. devices on patients, hospitals, doctors).Fourthly, the gateway should enable monitoring a patient’shealth status in real time and mediate appropriate services(e.g. diagnoses and prescriptions) from medical professionals.Finally, data privacy, security, validity and integrity are of highimportance when human life is at stake.

ZigBee- and Bluetooth-based wireless sensors have beenwidely used in BANs for context detection in health andwellness applications. A wireless sensor network developed inthe Codeblue [9] project provides routing, naming, discovery,

and security for wireless medical sensors, PDAs, PCs, andother devices that may be used to monitor and treat patients innetwork environments of various densities. The In-Home Mon-itoring [10] is a ubiquitous health care system for real-timemonitoring of patients’ locations with GPS and patients’ vitalsigns with wearable ZigBee sensors. Several systems [11], [12]have been developed to monitor electrical activity of a humanheart in real time with an ECG (electrocardiogram) sensor, anECG console and a ZigBee module. ECG data collected froma sensor network is transmitted to a server through a gateway.In many ubiquitous health care systems using Bluetooth, asensor node coupled with a Bluetooth module communicateswith a Bluetooth-enabled device. An example of this approachis demonstrated in the LifeGuard [13] which is vital signmonitoring system for astronauts. It uses a wearable wiredsensor kit that communicates with a Tablet PC via Bluetooth.In another example [14] a mobile phone acts as a gatewaywhich connects between sensor nodes and CDMA networkor other devices having Bluetooth support. Viswanathan etal. [15] proposed a distributed resource provision frameworkin which nearby computing devices (laptops, tablets, PDAs)use Bluetooth or ZigBee to collect and preprocess data beforesending them further. Today many ubiquitous health care sys-tems utilize Personal Health Devices (PHDs) which commu-nicate over protocols standardized in ISO/IEEE 11073. In [16]an integrated gateway architecture is proposed to collect datafrom various PHDs using a range of communication methodssuch as ZigBee, Bluetooth and USB.

As shown in the aforementioned examples, many ubiquitoushealth care systems use either ZigBee or Bluetooth technologyto send sensor data to a gateway. If a ubiquitous health caresystem is implemented using only ZigBee then transmittingdata requiring QoS (Quality of Service) could cause delaysdue to low transmission rates (ZigBee’s maximum trans-mission rate is 250kbps). Conversely, if a ubiquitous healthcare system is implemented using only Bluetooth, a largeamount of sensors could not be used unlike with ZigBee,and it would be harder to implement a monitoring systemsupporting multiple sensors that might be distributed overseveral targets (e.g. collective health monitoring of a group ofsoldiers). Viswanathan et al’s [15] idea of distributed resourceprovision is viable in well-defined contexts but when the useris mobile it becomes difficult to provide a dynamic architectureto support the framework. The integrated gateway in [16]solves this problem and it supports multiple data transmissionmethods (ZigBee, Bluetooth, USB) but it can be only usedwith ISO/IEEE 11073 standard compliant PHDs.

III. MOBILE GATEWAY FOR THE U-HEALTHCARE SYSTEM

Figure 2 presents the proposed mobile gateway concept forBANs in the context of ubiquitous health care. The mobilegateway, carried by the user, mediates sensor data between aBAN and a TCP/IP network. Each sensor node is responsibleof sensing particular type of data (e.g. ECG, heart rate, bodytemperature) and propagating data packets to a sink node overZigBee protocol. The sink node communicates with the mobile

140

gateway using a Bluetooth module. After receiving a datapacket, the mobile gateway parses and preprocesses the databefore sending it to a remote database. Database connection isestablished over the Internet via WiFi, 3G or 4G connection.The user can use additional health care services directly onthe mobile device or through a website. Furthermore, the useris able to contact doctors by SMS or voice call. The mobilegateway can also be programmed to analyze retrieved dataand send alarms to medical professionals. At the same time,medical professionals can monitor the user’s health remotelywith their mobile devices or through a web interface. Remotemonitoring is done almost in real time; only a small delay oc-curs due to parsing, preprocessing, transmission and retrievalof data. Data in a centralized database can be made availablefor third party users such as researchers. When storing healthdata into an online database special attention must be givento privacy protection. Data should be coded or anonymizedto avoid revealing users’ identities. Data encryption schemesshould be applied before the system is deployed to real use.

Fig. 2. Proposed mobile gateway concept

A. Design and implementation

In the mobile gateway design we emphasized adaptabilityand portability. As a result, the system can be adapted toany ZigBee-based wireless sensor network. Figure 3 illustratestechnical design of the system including modules and connec-tions between the modules. The gateway was implemented inJava and it can potentially run on any Java-enabled mobiledevice, but some modifications might be required to user in-terface according to the target device’s capabilities. Graphicaluser interface of the prototype gateway (Figure 4) is builtwith the Standard Widget Library (SWT) from the Eclipseproject. Remote database is implemented with MySQL. Fordeveloping and testing the system we use Nokia N810 devicesrunning on Linux. N810 represents old hardware today butit is powerful and versatile enough for our purpose. It has,among other features, a large screen with 800x480 resolution,physical QWERTY keyboard, GPS receiver, camera, WLANand Bluetooth.

Fig. 4. Mobile gateway’s monitoring screen visualizing ECG data

The U-Healthcare gateway on a mobile device uses twotypes of connections: Bluetooth and Internet (e.g. 4G, 3G,WLAN). We utilize Bluetooth for communication between themobile device and a ZigBee sensor node which acts as the sinknode of BAN. The sink node has a special Bluetooth moduleattached which enables communication with another Bluetoothdevice. The sink node sends captured data over Bluetooth tothe mobile gateway, and the mobile gateway in turn sendscontrol messages to the sink node when necessary. Bluetoothcommunication is explained in detail in the following section.Internet connection is needed for three purposes: 1. storingand retrieving data to/from remote database, 2. accessingthe server’s web-based interface for additional services (e.g.medical, support), and 3. connecting to another mobile device(e.g. doctor) for consultancy.

The internal gateway architecture consists of three mainmodules: Bluetooth Module, Data Module and User Inter-face Module (see Figure 3). Bluetooth Module takes careof discovering Bluetooth devices and services, connectingto a sensor network and sending control messages to andreceiving data from a ZigBee sink node. This process ishandled by the BTManager which runs dedicated componentsfor outgoing and incoming communication (BTSender andBTReceiver). These components interact with the local datamanager and the user interface. Data Module’s responsibilityis to store collected data first locally and then remotely to acentralized database over the Internet connection. DBManagertakes care of transferring local data to a remote database.DataManager acts as a mediator between BTReceiver andlocal data storage (SensorDataRepository). SensorDataRepos-itory consists of multiple SensorNode objects each of whichrepresents a physical sensor node such as EGC or temperature.A SensorNode stores locally SensorData objects which arecaptured from that particular node. We designed the Sen-sorNode component so that it can hold one or more sensordata values and it can be extended to meet the requirementsof a specific physical sensor node. This makes the systemadaptable for any kind of ZigBee sensors. SensorNode alsoprovides an event listening mechanism which allows othercomponents to be notified when a particular type of data is

141

Fig. 3. Technical design of the mobile gateway

updated. In our design, DBManager and Monitor componentsare listening for changes in the SensorNodes. Once data is inremote database, it can be accessed by any other application,such as a web monitoring system, with appropriate credentials.While SensorNodes store only recent data for fast retrieval,older data can be retrieved from remote database when needed.User interface module is responsible of receiving commandsfrom the user and plotting a real-time graph of sensor data(Monitor). The graph can also be traversed chronologically.The user can change preferences and properties of the gateway(e.g. database connection details), choose a Bluetooth devicefor connection (BTConfig), capture images and send controlmessages to sink node (Tools), and send an SMS or call doctor(Communications).

The database for storing the sensor data on the server isdivided into four tables (Figure 5). The main table is sen-sor data which holds all captured sensor data. Primary key ofthe table is sensor data id. Type id is a foreign key pointingto the sensor data type table which hosts the different typesof sensor data. This allows us to store any kind of sensordata in a centralized manner. Target id refers to the primarykey field on the target table which stores information of thetarget patient. Timestamp field of sensor data table denotesthe capture time of a sensed value. Location id foreign keyrefers to location table which stores physical location of thesituation where the data was sensed. In order to utilize GPS inthe future, the location table has latitude and longitude fields.

B. Sensor nodes and Bluetooth communicationWireless sensor nodes in our system comprise an AT-

mega128L MCU (Micro Control Unit), four internal sensors

Fig. 5. Database schema for the system

(temperature, humidity, light and infrared), an A/D converter,UART and ZigBee RF transceiver. Additionally, the sink nodehas a Bluetooth module which operates at 3 volts and uses theunlicensed 2.4GHz ISM band with a communication rangebelow 30 meters. External sensors such as ECG, heart rateand body temperature can be attached to sensor nodes bythe UART interface. Sensor nodes run TinyOS which is aopen source component-based event-driven operating systemfor wireless sensor networks. Figure 6 illustrates the structureof a sink node coupled with a Bluetooth module.

142

Fig. 6. Structure and protocol stack of a sensor node coupled with a Bluetooth module

The protocol stack of a sensor node comprises application,network, MAC and PHY layers as shown in Figure 6. In oursystem the application layer has four functions: CommandQuery analysis/response handles command queries receivedfrom other sensor nodes or the mobile gateway; Sensed DataCollection gathers information and sensed data from localsensors and adjacent sensor nodes; Data Analysis analyzessensed data; and Sensor Object decides the sensor node’s role(sink node, router node or source node). The network layerincludes two functions: Routing configuration enables a sourcenode to configure routes up to a sink node and to a routernode; and Sensed Data Forwarding enables a router node toforward data from a source node to a sink node. Finally, theMAC/PHY layers follow the functions defined in the IEEE802.15.4 standard for wireless personal area networks.

The data collected from a sensor network are transmittedto the mobile gateway through the following process. A sinknode collects captured data (e.g. vital signs) from adjacentnodes using the ZigBee protocol. The collected data arriveat application layer after passing through lower layers ofthe sink node. In the application layer, the data are packedaccording to a predefined data format as exemplified in Table I.The packet header consists of sequence number of the datapacket (Seq) and unique Node ID. The packet payload consistsof two-byte values of each sensor type, allowing a singlenode to send readings from multiple sensors at once. Createdpackets are transmitted to the Bluetooth module through theUART interface. The Bluetooth module forwards the packetsat maximum data rate 1Mbps to the mobile gateway.

A set of control messages were defined between the mobilegateway and sensor network for changing variables of the datacollection process such as sensing interval and starting/pausing

TABLE IAN EXAMPLE OF DATA PACKET FORMAT

Header Payload

2 bytes 1 byte 2 bytes 2 bytes 2 bytes

Seq Node ID Temp HR ECG

TABLE IICONTROL MESSAGES SENT FROM MOBILE GATEWAY TO THE SINK NODE

Code Value

Start all A (1B) 1 (1B)

Start B (1B) Node ID (1B)

Stop all A (1B) 0 (1B)

Stop C (1B) Node ID (1B)

Sensing period D (1B) Period (2B)

sensing (Table II). First column denotes the message type,the second column contains the operation codes, and the lastcolumn shows appropriate values and lengths of operations inbytes. The control messages are submitted through a mobiledevice’s Bluetooth transceiver to a sink node which delegatesmessages onward. By using the control messages, the mobilegateway can control sensor nodes by requesting an individualnode or all nodes to start/stop or to change sensing period.

In most cases, sensor networks having an aim to measureand monitor health status of a single person rely on singlehop communication. In order to transmit data over a longerdistance, a multi-hop network model is required. An example

143

of such situation is monitoring a group of geographicallydistributed targets using a single gateway. When multi-hopcommunication is used in a sensor network, routing protocolssuch as AODV (Ad-hoc On Demand Vector) and cluster treeschemes can be used to deliver captured data from a sourcenode to a sink node, and to manage and control the sensornetwork.

IV. DISCUSSION

The proposed mobile gateway can be used in variouscontexts where body area networks are utilized. For example,a personal trainer can monitor an athlete’s physical statuswith a mobile device through body sensors attached to theathlete’s body. Soldiers in war or training can wear a BANand medics can monitor their status’ on mobile devices col-lectively. A BAN can be also attached to a patient residingfar from medical facilities. In such cases, collected sensordata are transmitted to a hospital through the mobile gateway.Doctors and nurses in the hospital receive the data and canconstantly monitor the patient’s status. If the status changes,an immediately respond to the situation can be given. Finally,students in medicine, physical education, computer scienceand electrical engineering can utilize the proposed mobilegateway for educational purposes.

To investigate the proposed mobile gateway’s feasibility,both user and system evaluations are needed. Jovanov andMilenkovic [17] proposed success factors for BAN-basedubiquitous health care system in domains of user satisfactionand technical operation. User satisfaction is determined bywearability, ease of use, meaningful feedback, price, privacyand security. Technical aspects include low-power operation,seamless sensor integration, integration into other systems,user identification and localization. These success factorscould be the basis for an extensive system evaluation to proveusefulness of the proposed mobile gateway.

Collecting vital signs from a human body over time islikely to produce a large quantity of data to be managedsecurely and efficiently. In this paper we focused on thegateway implementation and used data communication with asingle server. The system was tested in laboratory conditionsonly, hence data management challenges of the real world areyet to be tackled. One promising solution to the problem ofmanaging large amounts of medical data is to use a cloudcomputing platform such as the ones presented in [18]–[20].When properly implemented using standardized technologies,the cloud computing approach allows large quantities of datato be collected, processed and accessed in distributed, efficientand secure manner.

One of the issues we did not cover in this study is thesystem’s energy efficiency that is a key factor in long-termusage. It has been shown in [21] that wireless technologieslike Bluetooth can fail to provide support for energy efficientsystems. Bluetooth 4.0 with low energy technology can bea solution to this as it provides significantly lower powerconsumption than older Bluetooth versions. Another potentiallow-energy wireless communication technology is the ANT

but it suffers from low bandwidth and lack of support in mobiledevices. More discussion of prevalent Body Area Networktechnologies is available in [22].

V. CONCLUSIONS AND FUTURE DEVELOPMENT

We have presented a mobile gateway and the U-Healthcaresystem to bridge the connection between a body area networkand the internet using Bluetooth. The requirements for suchsystem stem from the need of remote health care and well-ness services for monitoring in casual situations and in ruralcontexts. Light-weight wireless sensors enable such portablehealth care services. When enhanced with a mobile gateway,the system becomes truly ubiquitous as sensed data can becollected, processed and disseminated in real time regardlessof location and time. The proposed mobile gateway allowsa patient to observe real time or past vital signs though agraphical monitor, to control the sensor nodes and to com-municate with medical professionals. Bluetooth enables fasterdata rate in comparison to ZigBee protocol, thus large amountsof sensed data can be transferred between sink nodes and themobile gateway. As the proposed system integrates Bluetoothwith a ZigBee-based BAN, it overcomes the problems inprevious ubiquitous health care systems regarding single-pointdata aggregation (only Bluetooth) and slow transmission froma sink node to a gateway (only ZigBee). Furthermore, flexibledesign of the proposed gateway enables its adaptation to anysituation where ZigBee-based BANs require a mobile gateway.

As future work we will evaluate the performance of theproposed middleware by collecting data from real test subjects.We will also improve the system by implementing a server-based analysis component which will analyze sensor datain order to provide automatic alarms and diagnoses. GPSnavigation system on the mobile device will be utilized foruser localization. In BAN, we seek to investigate cross-layerdesign for reducing energy consumption and genetic algo-rithms for node deployment techniques and routing protocols.Multi-hop communication in ZigBee network will also be animportant issue if the system is to be deployed to physicallylarger contexts. We aim to investigate the gateway’s usage inother fields such as professional sports and military wheredifferent BANs are used. We also seek to experiment with thenew Bluetooth low energy protocol to achieve higher energyefficiency. Finally, privacy concerns must be solved and ethicalissues of patient monitoring discussed before the system canbe deployed in real environment.

ACKNOWLEDGMENT

We would like to express our gratitude to Sejin Oh forcontributing to the development of the ZigBee sensor networkand Peter Boda at the Nokia Research Centre in Palo Alto(now Sunnyvale), USA, for kindly donating N810 devicesfor this project. This work was supported by the new facultyresearch fund of Ajou University (2012).

144

REFERENCES

[1] G.-J. Hwang, T.-C. Yang, C.-C. Tsai, and S. J. Yang, “Acontext-aware ubiquitous learning environment for conductingcomplex science experiments,” Computers & Education, vol. 53,no. 2, pp. 402 – 413, 2009. [Online]. Available:http://www.sciencedirect.com/science/article/pii/S0360131509000669

[2] H. Ogata and Y. Yano, “Context-aware support for computer-supportedubiquitous learning,” in Proceedings of the 2nd IEEE InternationalWorkshop on Wireless and Mobile Technologies in Education, 2004,pp. 27–34.

[3] M. Adkins, J. Kruse, and R. Younger, “Ubiquitous computing: Om-nipresent technology in support of network centric warfare,” in HICSS,2002, p. 40.

[4] M.-S. Ju and S.-S. Kim, “A study on the use of ubiquitous technologiesin military sector,” International Journal of Contents, vol. 2, no. 2, pp.6–9, 2006.

[5] P. Blythe, “Rfid for road tolling, road-use pricing and vehicle accesscontrol,” in RFID Technology (Ref. No. 1999/123), IEE Colloquium on,1999, pp. 8/1–816.

[6] M. Karkkainen, “Increasing efficiency in the supply chain for shortshelf life goods using rfid tagging,” International Journal of Retail &Distribution Management, vol. 10, no. 31, pp. 529–536, 2003.

[7] D. Bujan, D. Martin, O. Torices, D. Lopez-de Ipina, C. Lamsfus,J. Abaitua, and A. Alzua-Sorzabal, “Context management platform fortourism applications,” Sensors, vol. 13, no. 7, pp. 8060–8078, 2013.[Online]. Available: http://www.mdpi.com/1424-8220/13/7/8060

[8] Y.-P. Huang, T.-W. Chang, W.-P. Chuang, and F. Sandnes, “An interactivehandheld device-based guide system using innovative techniques,” inInnovative Computing, Information and Control, 2007. ICICIC ’07.Second International Conference on, 2007, pp. 217–217.

[9] D. Malan, F. J. Thaddeus, M. Welsh, and S. Moulton, “CodeBlue: An Ad Hoc Sensor Network Infrastructure for Emergency MedicalCare,” in Proceeding on the MobiSys 2004 Workshop on Applicationsof Mobile Embedded Systems. ACM, 2004, pp. 12–14.

[10] G. Virone, A. Wood, L. Selavo, Q. Cao, L. Fang, T. Doan, Z. He,R. Stoleru, S. Lin, and J. A. Stankovic, “An assisted living orientedinformation system based on a residential wireless sensor network,” inProceedings of the IEEE 1st Conf. on Distributed Diagnosis and HomeHealthcare (D2H2 06), pp 95100, Apr 2006. 7. BIBLIOGRAPHY 219,2006, pp. 2–4.

[11] B. Kim, Y. Kim, I. Lee, and I. You, “Design and implementation of aubiquitous ecg monitoring system using sip and the zigbee network,”in Future Generation Communication and Networking (FGCN 2007),vol. 2, 2007, pp. 599–604.

[12] S.-J. Oh and C.-W. Lee, “U-healthcare sensorgrid gateway for connect-ing wireless sensor network and grid network,” in Advanced Commu-nication Technology, 2008. ICACT 2008. 10th International Conferenceon, vol. 1, 2008, pp. 827–831.

[13] K. Montgomery, C. Mundt, G. Thonier, A. Tellier, U. Udoh, V. Barker,R. Ricks, L. Giovangrandi, P. Davies, Y. Cagle, J. Swain, J. Hines, andG. Kovacs, “Lifeguard - a personal physiological monitor for extremeenvironments,” in Engineering in Medicine and Biology Society, 2004.IEMBS ’04. 26th Annual International Conference of the IEEE, vol. 1,2004, pp. 2192–2195.

[14] L. Zhong, M. Sinclair, and R. Bittner, “A phone-centered body sen-sor network platform cost, energy efficiency amp; user interface,” inWearable and Implantable Body Sensor Networks, 2006. BSN 2006.International Workshop on, 2006, pp. 4 pp.–182.

[15] H. Viswanathan, B. Chen, and D. Pompili, “Research challenges incomputation, communication, and context awareness for ubiquitoushealthcare,” Communications Magazine, IEEE, vol. 50, no. 5, pp. 92–99,2012.

[16] K.-H. Park and J.-G. Pak, “An integrated gateway for various phds inu-healthcare environments,” Journal of Biomedicine and Biotechnology,vol. 2012, 2012.

[17] E. Jovanov and A. Milenkovic, “Body area networks for ubiquitoushealthcare applications: Opportunities and challenges,” Journal ofMedical Systems, vol. 35, no. 5, pp. 1245–1254, 2011. [Online].Available: http://dx.doi.org/10.1007/s10916-011-9661-x

[18] C. Doukas and I. Maglogiannis, “Bringing iot and cloud computing to-wards pervasive healthcare,” in Innovative Mobile and Internet Servicesin Ubiquitous Computing (IMIS), 2012 Sixth International Conferenceon, July 2012, pp. 922–926.

[19] Z. Rashid, U. Farooq, J.-K. Jang, and S.-H. Park, “Cloud computingaware ubiquitous health care system,” in E-Health and BioengineeringConference (EHB), 2011, Nov 2011, pp. 1–4.

[20] Y.-Y. Ou, P.-Y. Shih, Y.-H. Chin, T.-W. Kuan, J.-F. Wang, and S.-H. Shih,“Framework of ubiquitous healthcare system based on cloud computingfor elderly living,” in Signal and Information Processing AssociationAnnual Summit and Conference (APSIPA), 2013 Asia-Pacific, Oct 2013,pp. 1–4.

[21] D. Bhatia, L. Estevez, and S. Rao, “Energy efficient contextual sensingfor elderly care,” in Proceedings of the 29th Annual InternationalConference of the IEEE EMBS, pp 95100, Apr 2006. 7. BIBLIOGRAPHY219, 2006, pp. 2–4.

[22] M. Patel and J. Wang, “Applications, challenges, and prospective inemerging body area networking technologies,” Wireless Communica-tions, IEEE, vol. 17, no. 1, pp. 80–88, 2010.

145