interference study of medical body area network for ... · interference study of medical body area...

Interference Study of Medical Body Area Network for Proactive Performance Improvement 23

Interference Study of Medical Body AreaNetwork for Proactive Performance

Improvement

Gengfa Fang1 , Eryk Dutkiewicz2 , and Rein Vesilos3 , Non-members

ABSTRACT

Medical Body Area Network (MBAN) is a newwireless communications technology designed to sensehuman’s vital signals through tiny nodes in, on andaround the human body wirelessly. MBAN will playan important role in enabling ubiquitous and non-invasive telemetry and healthcare systems in the fu-ture. In this paper, we analyze the interference inMBAN both from legacy wireless networks and neigh-bouring MBANs. We then study the possible solu-tions to fight against these interferences in MBANincluding the channel sharing scheme to solve theinter-network interference. Experiments are carriedout based on MQWIN400 radio platform to studythe wireless channel in 400MHz band. We found thereceiver’s mobility makes the channel quite dynamicbecause of multipath effect.

Keywords: Medical Body Area Networks, Wireless,Interference

1. INTRODUCTION

The high aging population is becoming one of thebig issues for developed countries like Australia. Anincreasing size of the aging population increases thecost of healthcare and is a significant challenge forthe governments, healthcare providers and healthcareindustry. Medical Body Area Network (MBAN) [4]consists of medical devices that communicate in andaround the human body. MBAN can provide con-tinuous and unobtrusive sensing and monitoring ofhealth parameters in and on human bodies. For ex-ample, wireless medical devices implanted in the hu-man body (Pill Camera) [4] offer an unobtrusive andsafe method for continuous health monitoring and areexpected to enable controlled drug delivery in the fu-ture. Unobtrusive and continuous parameter moni-toring in natural physiological conditions opens theprospect of more efficient diagnostic methods.

MBANs can monitor vital body signs such asheart-rate, temperature, blood pressure, ECG, EEGand pH level of patients. By replacing cables with

Manuscript received on July 31, 2011 ; revised on November1, 2011.1,2,3 The authors are with Department of Electronic En-

gineering Macquarie University, NSW, Australia. , E-mail:[email protected], [email protected] [email protected]

wireless links, MBANs can provide less invasive andmore comfortable and efficient systems both in hos-pital and outside the hospital. MBANs are of partic-ular interest to the healthcare sector to provide effi-cient healthcare services and ongoing clinical manage-ment. Examples of implantable devices that can beequipped with wireless transceivers include cochlearimplants, retina implants, glaucoma sensors, intracra-nial pressure sensors, capsule endoscopes, glucose sen-sors, insulin pumps, heart pace makers, cardiac moni-tors, deep brain stimulators, visual neuro-stimulators,and many more.

In addition to medical application, the technologyhas important veterinary implications and its devel-opment is therefore crucial for almost all bioscientificendeavours. MBANs will play an important role inenabling ubiquitous communications, creating a hugepotential market. In the area of healthcare, accord-ing to the World Health Organization’s statistics, mil-lions of people suffer from obesity or chronic diseasesevery day, while the aging population is becoming asignificant problem.

In this paper, we analyze the interference issue inMBAN both from legacy wireless networks and neigh-boring MBANs. We then propose truthful onlinechannel sharing based solution that distribute chan-nel efficiently and truthfully by punishing MBANs formisreporting their channel requests including time,duration and bid for the channel. MQWIN400(Macquarie University Wireless Implantable Node at400MHz) platform is introduced and the 400MHzMICS band channel measurements are also providedconsidering the effect of the human tissue using a spe-cialized solid human body phantom.

The remainder of the paper is organized as follows:Section II analyzes new opportunities of MBAN bydescribing new applications that MBAN can support.In section III, we analyze the interference in MBANand its possible solutions. MQWIN400 platform andrelated research results are provided in Section IV,followed by the conclusion in Section V.

2. MBAN OPPORTUNITIES

The opportunities of MBANs come from the newapplications that MBANs can enable. MBANs canprovide ubiquitous monitoring for healthcare. Unlikeconventional bedside patient monitoring, MBANs canprovide a point of non-invasive care systems to pa-

24 ECTI TRANSACTIONS ON COMPUTER AND INFORMATION TECHNOLOGY VOL.6, NO.1 May 2012

tients, the elderly, and infants in both hospital-basedand home-based scenarios. Monitoring, autonomousdiagnostic, alarm, and emergency services, as well asmanagement of electronic patient record databasescan all be integrated into one system to better servepeople. MBAN can continuously capture quantita-tive data from a variety of sensors for longer peri-ods. By addressing challenges such as the energyand throughput tradeoff, MBANs will enable tele-health applications because of their human-centricityand will facilitate highly personalized and individualcare. As Figure 1 illustrates, MBANs integrated witha higher-level infrastructure will likely excel in health-care scenarios, serving the interests of multiple stake-holders. In addition to delay-insensitive applicationssuch as longitudinal assessment, MBANs that can of-fer real-time sensing, processing, and control will aug-ment and preserve body functions and human life.MBANs researchers are already working to improvedeep brain stimulation, heart regulation, drug deliv-ery, and prosthetic actuation. MBANs technologywill also help protect those exposed to potentiallylife-threatening environments, such as soldiers, firstresponders and deep-sea and space explorers.

Fig.1: MBAN and its Application Architecture.

Totally implantable devices are less susceptible toinfection. MBANs will enable the development ofnew medical applications that are crucial to sustain-ing affordable high quality health service delivery andhealth care in the increasingly aging population. Theresearch into MBAN will lead to safe, less intru-sive, reliable and cost effective solutions for a widerange of ambulatory and intensive care applications.It will enable novel medical and laboratory investi-gation systems, tailored post-operative drug deliverysystems, artificial organs (eg pancreas) and precisionmicro-surgery.

More recent developments in the area of medicalwireless technology include the formation of the Con-tinua Alliance [6], an open industry coalition of over200 companies with the goal of improving health-care via the promotion of interoperability and con-nected health care technology and solutions and theformation of the European Telecommunications Stan-dards Institute (ETSI) eHealth group [7] to coor-dinate ETSI activities in this area. The FCC cre-ated the Medical Device Radiocommunication Ser-vice (MedRadio Service) [7] incorporating the exist-

ing MICS band and including the adjacent “wing”bands at 401-402 MHz and 405-406 MHz. Altogether,the MedRadio Service will provide a total of fivemegahertz of contiguous spectrum on a secondary ba-sis and non-interference basis for advanced wirelessmedical radiocommunication devices used for diag-nostic and therapeutic purposes in humans. In addi-tion, standardisation activities are under way in IEEE[9] for the physical and medium access control mech-anisms for MBAN applications including implantablemedical devices. The existence of a MBAN stan-dard will provide opportunities to expand these prod-uct features, better healthcare and wellbeing for theusers. It will therefore result in economic opportunityfor technology component suppliers and equipmentmanufacturers.

3. WIRELESS CHANNEL AND CHAL-LENGES IN MBAN

3.1 MBAN wireless channel and challenges

An important requirement in MBAN is the energyefficiency of the system. A medium access control(MAC) layer is the most suitable level to address en-ergy efficiency. This layer is used to coordinate nodeaccess to the shared wireless medium. MAC is thecore of any communication protocol stack whose per-formance provides the basis for achieving Quality ofService (QoS) in any wireless networks. A versatileMAC should support distinct applications and differ-ent types of data such as continuous, periodic, burstand nonperiodic along with high level QoS. MACplays a major determining factor in improving overallnetworks performance. The fundamental task in theMAC protocol is to avoid collisions and to prevent si-multaneous transmissions while preserving maximumthroughput, minimum latency, communication relia-bility and maximum energy efficiency.

There are lots of other problems to be solved toprovide an overall solution of MBAN MAC, for exam-ple power efficiency, power control and inter-networkinterference coordination. In order to design an ex-tremely power efficient MBAN MAC, the uniquenessof the MBAN network topology and communicationcharacteristics need to be utilized. Unlike other IEEE802.15.4 [10] based sensor networks, MBAN is basedentirely on a star topology and the MBAN coordi-nator usually has enough power and computing ca-pacity. In addition, most of the communication isfrom MBAN nodes to the MBAN coordinator, i.e.uplink and downlink traffic is asymmetric. These al-low to design a receiver (i.e., coordinator) initiatedMBAN MAC protocol, so that MBAN nodes will bein a power saving state unless they are activated bythe coordinator to wake up for data exchange. In-terference coordination and network co-existence arekey issues for the performance of MBAN. PreviousBayesian game based solutions [6] [7] are unrealis-tic in that they assume a symmetric case where full


knowledge of channel state information of interfer-ing channels and number of interfering networks areavailable. A practical MAC power control and co-existence protocol based on IEEE 802.15.6 standardis left to be defined to coordinate interfering MBANsto control transmission power and share frequencybands to minimize the effect of inter-network inter-ference.

Zarlink’s ZL70101 and ZL70102 [1] are the worldlatest transceiver chips for medical implantable ap-plications, which provide a second wake-up radio toactivate the primary transceiver so as to save powerby turning off primary transceiver but how MAC canutilize this to achieve ultra low power communicationand meet the QoS requirements of medical applica-tions is an open issue even for the IEEE 802.15.6 stan-dard. [11]-[13] studied the channel state of frequencybands for on-body and in-body MBAN and concludedthat there are channel diversity among different com-munication links both in the frequency and time do-mains and proper MAC mechanisms are necessaryto protect one transmission from the interference ofother nearby concurrent transmissions, like Earth Ex-ploration Satellite Service (EESS), Meteorological-Satellite Systems (MetSat) and Emergency PositionIndicator Radio Beacon (EPIRB) [5] in the same fre-quency or neighbour band. Thus a MBAN MAC de-sign can explore user diversity among same sourcelinks and different source links to improve frequencyefficiency and reliability.

Deep fading occurs when a signal received by awireless node is highly degraded due to environmentalfactors such as temporal physical obstructions. Sig-nals can also be degraded due to interference fromother wireless networks sharing the same radio spec-trum as in Figure 2, where an external system sendsa strong signal in the MICS band. These phenom-ena may result in the inability of the wireless nodeto maintain uninterrupted communication with othernodes leading to degraded performance in terms of re-liability and energy efficiency. Combating deep fadingand interference is particularly significant in appli-cations where reliable uninterrupted communicationis required, such as in real-time monitoring of vitalsigns in an intensive care unit. Recent investigationsof MBANs identified time-varying fading of radio sig-nals in the human body as the major problem forreliable low power operation [6].

3.2 Interference from Nearby MBANs

Unlike cellular networks, MBANs are randomlydistributed networks where two or more MBANs mayoverlap with each other and interfere with each otherdue to the limited available frequency bands. For ex-ample in a crowded bus more there 10 people maysit close to each other, so that each person’s MBANwill interfere with the others. Severe interference willdecrease the SINR dramatically and as a result cause

Fig.2: MICS band is not clean in Australia.

throughput degradation and packet loss. Packet lossalso leads to energy waste and decrease in energy ef-ficiency of MBAN nodes. Since stability is a criticalissue in MBANs, interference may cause life criticalpackets loss and hence is a threat to patients’ liveswhen MBANs are used in the healthcare sector.

To mitigate inter-network interference, many in-terference mitigation and coordination technologieshave been developed to allow multiple networks exist.However, as shown by [16][17], when inter-network in-terference becomes too severe, channel sharing amongnearby networks becomes more efficient than anyother interference mitigation and coordination tech-nology in practice. As depicted by Figure 3, i.e., wehave

log(h11

n0)t1+log(

h22

n0)>(log(

h11

h21+n0)+log(

h22

h12 + n0))(t1+t2)

where hij is the channel gain between transmitteri and receiver j, and and hij is the intra-network gainif i = j and inter-network interference gain for i = j.

Fig.3: MBAN An example of channel sharing.

Channel sharing among MBANs introduces signif-icant design challenges because of the mobility andsafety requirements of MBANs. First, the channelsharing scheme must be on demand, so that MBANscan request for the channel access of each frame. Sec-ond, the channel access scheme has to make decisionson-the-fly, i.e., without the knowledge of future eventson which networks will subsequently join or leave the


channel sharing game. Third, channel sharing mustbe self enforcing so that each MBAN will benefit fromsharing the channel with other MBANs instead of op-erating uncooperatively and causing interference toothers. Forth, it must be truthfulness enforcing sothat no MBANs can gain unfair advantage by ma-nipulating their channel requests.

Taking the above into consideration, we model thechannel sharing problem as an online channel accessauction, where users (i.e., MBANs) will request forchannel access whenever channel bandwidth is neededto support the communications in the networks. Eachrequest consists of a monetary bid, and channel accessinformation which includes start time, end time andchannel access length in the channel sharing problem.One of the users works as the auctioneer, which pro-cesses the requests on-the-fly, and forward the chan-nel access grants to all the users. By decoding thechannel access grants, each user can access the chan-nel that is granted to him. In this way, users canrequest and share the channel with its neighbours towithout collision or severe inter-network interference.

Although the online channel access auction designcan easily meet the requirements of online real-timerequirements, it opens up vulnerabilities to selfishusers that try to gain unfair advantages over others bymanipulating their requests, for example, a user canfalsely report its channel request in terms of channelaccess duration, start time and end time, or bid toget more channel bandwidth at low cost. This chan-nel access manipulation can even block other usersfrom transmission at all if it is not solved properly.Therefor, an efficient online channel access auctiondesign for MBANs’ coexistence needs to address thisproblem so that misbehaved users will be punished;as a result, channel bandwidth can be fairly sharedamong users.

Assume that there areN users (i.e., MBANs) shar-ing a unique channel frame by frame, and each frameis identical for all the users. A central authority,for example, one of the MBANs, will decide the al-location of each frame to n users. Upon detectinga need for channel bandwidth from the applicationlayer, users send requests to the central authority forchannel access. Each request of user i is character-ized by vi = (ai, di, li, wi). We refer to ai and di asthe arrival time and deadline respectively, and referto li and wi as channel access length and bid accord-ingly. Time is slotted (frame by frame) and all re-quests have channel access length between [1,∆], i.e.li ∈ [1,∆] indicating number of frames required. Wehave ai ≤ di, 0 < li ≤ (di−ai) and 0 ≤ wi < ∞. Eachuser can participate in the channel sharing procedureby simply sending its request vi. Upon receiving therequests from users, the central authority allocateseach user a subset of the frames so that users are as-signed disjoint subsets, during which the end nodesand coordinator in that MBAN can exchange packets

according to the IEEE 802.15.6 standard.

We model the above channel sharing problem asan online auction where each user (bidder) submit re-quests vi to the auctioneer (central authority) when-ever it needs channel bandwidth in terms of frames.The online auction procedure is randomly triggeredby new request arrival or current winner departure.Online channel sharing scheme needs to make deci-sions on-the-fly, i.e. without the knowledge of thefuture requests coming. As a result, the overall per-formance of the online channel sharing can be largelydegraded if users deliberately try to manipulate theirrequests. We use v = (v1, v2, . . . , vn) to denote therequests from n users for current frame t. Onlinechannel sharing scheme includes a channel alloca-tion rule q and a payment rule p, where q(v) =(q1, q2, . . . , qn) represents the channel allocation re-sult, p(v) = (p1, p2, . . . , pn) represents the amountuser i needs to pay, and qi ∈ {0, 1}, 0 ≤ pi < ∞. Forsimplicity, we consider one channel in this paper, sowe have

∑ni ai = 1. We will assume that users have

quasi-linear utility function, so the utility of user i forallocation qi(v) and pi(v) is qi(v)wi − pi(v).

The flexibility of the online auction method on theother hand makes the channel sharing design diffi-cult and challenging. Selfish and malicious users canutilize this flexibility to manipulate their requests tocontrol the auction outcome so as to have unfair ad-vantage over others. In online channel sharing, userscan cheat by not only rigging their channel accesslength and their bids for accessing the channel, butalso by falsely reporting their start time and dead-line for accessing the channel. A good online channelaccessing design needs to resist these selfish and mali-cious behaviors. One well-known solution is to makethe online auction truthful, i.e., no one can improvehis utility by misreporting his request.

We assume that users are selfish; as a result userwill not report channel request that has start-timeearlier than its true start-time and end-time laterthan its true end-time. This is a practical assumptionand makes sense for MBAN channel sharing problem,because each user will not benefit if it reports thechannel usage earlier than it really need and delayingits channel usage will result in disadvantage. Sinceeach user is selfish, it will only misreport its channelrequest that is no early start-time and no late end-time.

If we use v′

i = (a′

i, d′

i, l′

i, w′

i) to denote the falserequest and vi = (ai, di, li, wi) as the true request, we

have a′

i > ai, d′

i < di, w′

i < wi, and l′

i > li. We areinterested in designing truthful online channel sharingscheme that is, for every v = (v1, v2, . . . , vn) with

vi = (ai, di, li, wi) and every v′

i = (a′

i, d′

i, l′

i, w′

i), we

have qi(v)wi−pi(v) ≥ qi(v′

i, v(− i))w′

i−pi(v′

i, v(− i)),where v(− i) = (v1, . . . , v(i−1), v(i+1), . . . , vn), i.e.,the utility of user i is maximized if and only if shesubmits her request for the channel access truthfully.


We consider the unit-length case where each timea user requests only one frame, i.e. li = 1. We canthen simplify the channel access request by defining itas vi = (ai, di, wi). In the online channel sharing pro-cess, each user submits its request to the auctioneerfor channel access whenever it finds that its MBANneeds channel bandwidth for the communication be-tween its coordinator and the MBAN end devices. Asa result, the channel allocation decision talks place atthe beginning of each criticalframe, when new usersrequests arrive for channel bandwidth, or when a pre-vious winner finishes its usage of the channel.

Driven by the monotonic allocation methodologyabove, we apply the following channel allocation ruleq (HIGHEST BID FIRST) to allocate frames to users:for each critical frame t, allocate frame t to the useri with the pending request that meets the following:wi = max{wj : aj ≤ t ≤ dj}, i.e., the highest bid userfor current frame will be served first.

In order to be truthful implementable, the pricingscheme has to be bid independent and monotonic.Otherwise, users can increase its utility by cheating.In online channel auction, a user’s price when winningthe auction depends on other pending users and theirbids.Since the set of pending users depends on the time,we need to design a pricing scheme that can removethe time dependency and still be monotonic. Thismeans the price charged when a user cheats is no lessthan that when she reports truthfully. Taking theabove into account, we define the payment rule p as:

pi(v) = qi(v)wi −∫ wi

0

qi((ai, di, x), v−i)dx (1)

In words, a user will only have to pay the smallestvalue it could have reported in order to receive anallocation.

4. MQWIN400 BASED PRACTICES

In this section, we describe the MQWIN400, aMICS band hardware platform for the implantableMBAN research and the channel measurement activ-ities and results.

4.1 MQWIN400 Hardware and Network Ar-chitecture

As shown in Figure 4, MQWIN400 is based onTI’s SoC CC111[17], which is a low power sub-1GHz designed for low power wireless applications.It combines the excellent performance of the state-of-the-art RF transceiver CC1101 with an industry-standard enhanced 8051 MCU, up to 32 kB of in-system programmable flash memory and up to 4 kB ofRAM, and many other powerful features. The small6x6 mm transceiver package makes it very suitedfor applications with size limitations. MQWIN400

provides general purpose I/O interfaces to supportsensors like gyroscope, ECG, EMG, and other sen-sors. MQWIN uses the latest MICS band antennaANT-403-SP Splatch[18]. The Splatch antenna usesa grounded-line technique to achieve outstanding per-formance from a tiny surface mount element. Thisunique antenna is designed for hand- or reflow mount-ing directly to a product’s circuit board. Its low costmakes it ideal for volume applications. Unlike manycompact antennas, the Splatch exhibits good perfor-mance in proximity to objects and persons. Whilesuitable for a wide range of RF applications, the 403version was designed to cover the 402-405 MHz MICSband.

Fig.4: MQWIN400 platform.

A typical MQWIN400 network as shown in Figure5 consists of multiple implantable nodes in a patientand a coordinator node, which typically works as arelay that has both an MBAN radio such as IEEE802.15.6 and an 802.15.4 transceiver, to exchangecommunications between MBAN and the externalworld. Sensors collect physiological data from a pa-tient’s body, after that these data are transformed todigital signals and transmitted out to a gateway de-vice. It collects all the data from the WBAN nodesand forwards it to WAS. WAS will process the datareceived according to their types and IDs and thendisplay and save them. Network management packetscarry network information and will be used for net-work setup and provide node information on WAS.Each node can have multiple sensors. Each sensormay be treated as one data channel. WAS gets allthe data for all the sensors from WBAN through aUSB interface. The information in each packet cancorrespond to sensing data such as CO2, Oxygen andHP levels as well as network working information suchas Node ID, Channel ID and network state.

4.2 MICS band Channel Measurement andResults

In this subsection, we describe MICS band chan-nel measurements and initial results based on usinga solid human body phantom. Different types of hu-man tissue have different dielectric properties result-


Fig.5: MQWIN400 platform.

ing in different radio signal propagation effects. Radiopropagation is also affected by body posture, bodymovement and the amount of tissue in the radio path.Channel models also need to incorporate the antennatechnologies. Due to the relatively low frequency ofoperation in the vicinity of 400 MHz and a short op-erating distance, near-field effects are expected to besignificant.

In our experiments, we place the MBAN node in-side a human phantom which was designed to havethe same dielectric properties as the human muscle.A receiver is placed outside of the phantom and con-nected to a PC to save the RSSI information intoa file for processing. The implant wireless channelcharacteristics will be captured through a compre-hensive measurements campaign in typical usage en-vironments as shown in Figure 6. A MQWIN400 nodewill send 250 packets per second to the receiver. ATEach time a packet is received, the RSSI will be mea-sured and stored on the PC for processing.

We considered two scenarios so far. In the firstscenario, the distance between the transmitter andreceiver are fixed. There are two people workingaround the human phantom. We want to see the mul-tipath effects from the people nearby and the pathloss caused by the human phantom. The transmittedpower is 1mw and the distance between the implantnode and the external receiver is 1 meter. As shown inFigure 7 and 8, the average received RSSI is -69dBm.Activities of the people working around the phantomintroduce dBm dynamics to the pass loss.

Fig.6: Channel Measurement Setting.

In the second dynamic scenario in Figure 9 and 10,

we try to observe the received signal strength whenthe receiver node is placed on a person’s hand andthe person is moving. The average distance betweenthe transmitter and receiver is around 80 cm. Werun the experiment for 600 seconds. Figure 8 depictsthat the received signal level is quite dynamic vary-ing from -80dBm to -30dBm and has an average of-65dBm. The signal can go from good to bad in a veryshort time and may cause significant packet loss. Wecan see from Figure 8 that around 15% of the pack-ets have the received RSSI below -70dBm; as a re-sult, the MAC layer needs to be carefully designed tomeet these situations by adaptive power control andretransmission schemes.

5. CONCLUSIONS

In this paper we have focused on the understandingof interference in MBAN in terms of opportunities,challenges, as well as describing our research activi-ties and experiments. We firstly analyzed the poten-tial applications followed by the challenges of MBANfrom the point of energy efficiency, interference mit-igation and high data rate and network co-existenceissues. The experimental implant-node MQWIN400hardware was then described together with MBANnetwork. These were then used to set up experi-ments for implantable MICS band channel measure-ment. We explored two scenarios to see the pathloss caused by the human phantom and the effectof human movement. We found the receiver’s mobil-ity makes the channel quite dynamic, so the MAClayer needs to be designed to include channel awareschemes to handle these situations. Our future workwill be to further analyze the wireless channel by con-sidering more typical application scenarios and thenstudy network interference from nearby MBANs. Wewill then design a MAC protocol to deal with issues.

Fig.7: RSSI Measured in Static Scenario.


Fig.8: RSSI Distribution in Static Scenario.

Fig.9: RSSI Measured in Dynamic Scenario.

Fig.10: RSSI Distribution in Dynamic Scenario.

References

[1] Zarlink, “Medical Implantable RFTransceiver: Short Form Data Sheet,”http://www.zarlink.com.

[2] D.Miniutti, etc, “NarroMBANd Channel Char-acterization for Body Area Networks,” IEEE15-08-0421-00-0006-narroMBANd-channel-characterization-for-BAN.

[3] JhihCiangCai, ShihHeng, Cheng and ChingYao,Huang, “MAC Channel for MBAN-A Walk-ing Model Example,” IEEE 802.15-09-0562-00-0006.

[4] IEEE P802.15 Wireless Personal Area Net-works, “Channel Model for Body Area Network(BAN),” IEEE P802.15-08-0780-12-0006.

[5] Lubeck, ”Coexistence between Ultra Low PowerActive Medical Implants Devices and ExistingRadio Communication Systems and Services inthe Frequency band 401-402 MHz and 405-406MHz,” September 2006.

[6] H. Lee, H. Kwon, A. Motskin, and L. Guibas,“Interference-Aware MAC Protocol for WirelessNetworks by a Game-Theoretic Approach,” InProceedings of IEEE INFOCOM’09, 2009.

[7] R. Etkin, A. P. Parekh and D.Tse, “SpectrumSharing in Unlicensed Bands,” IEEE Journal onSelected Areas of Communication, vol. 25, no. 3,pp. 517–528, April 2007

[8] International Telecommunication Union/ITURadiocommunication Sector, ITU-R SA.1346”Sharing between the Meteorlogical Aids Serviceand Medical Implant Communication Systems(MICS) Operating in the Mobile Service in theFrequency 401-406 MHz,” Jan 1, 1998.

[9] ETSI EN 301 839-1 “Electromagnetic com-patibility and Radio spectrum Matters(ERM);Radio equipment in the frequencyrange 402 MHz to 405 MHz for Ultra LowPower Active Medical Implants and Accessories;Part 1: Technical characteristics, includingelectromagnetic compatibility requirements, andtest methods.”, European TelecommunicationsStandards Institute, 2002.

[10] Bradley P.D, “An Ultra Low Power, High Perfor-mance Medical Implant Communication System(MICS) Transceiver for Implantable Devices,”IEEE Biomedical Circuits and Systems Confer-ence, London, Nov, 2006.

[11] Continnua alliance:http://www.continuaalliance.org/about-the-alliance.html (accessed September. 2010).

[12] ETSI eHealth: http://www.etsi.org/website/technologies/eHealth.aspx (accessed September.2010).

[13] ET Docket No. 08-59 - Ammendment of theCommission’s Rules to Provide Spectrum for theOperation of Medical Body Area Networks andFCC 09-57, 29-Jun-2009.


[14] http://www.ieee802.org/15/pub/TG6.html (ac-cessed September. 2010)

[15] D. M. Davenport, B. Deb, and F. J. Ross,“Wireless Propagation and Coexistence of Med-ical Body Sensor Networks for Ambulatory Pa-tient Monitoring,” Body Sensor Networks (BSN2009), Berkeley, CA, June, 2009.

[16] K. Y. Yazdandoost and K. Sayrafian-Pour,“Channel Model for Body Area Network(BAN),” IEEE P802.15-08-0780-09-0006, April,2009.

[17] L. Matekovits1, M. Heimlich, and K. Esselle,“Numerical analysis of 2D tunable HIS on GaAssupport,” PIERS 2009.

[18] M. Heimlich, J. Kirshman, “Real-time Concur-rent Design of Adaptive High Frequency Cir-cuits,” IEEE Conference on Wireless and Mi-crowave Technology, April 2005.

[19] K. Malasri, K., L. Wang, “Securing wireless im-plantable devices for healthcare: Ideas and chal-lenges,” IEEE Communications Magazine, Vol.47, No. 7, pp. 74–80, 2009.

[20] C. A. Otto, E. Jovanov, A. Milenkovic, “AWBAN-based System for Health Monitoring atHome,” 3rd IEEE/EMBS International SummerSchool on Medical Devices and Biosensors 2006,pp. 20–23, 2006.

[21] G. Fang, E. Dutkiewicz, “BodyMAC: Energy Ef-ficient TDMA-based MAC Protocol for Wire-less Body Area Networks,” ISCIT 2009, 28-30September 2009.

Gengfa Fang received the Ph.D. de-gree in 2007 from the Institute of Com-puting Technology, Chinese Academy ofSciences, Beijing, China. He then joinedthe Canberra Research Lab of NationalICT Australia (NICTA) in 2007 in Can-berra, where he worked as a Researcher.In 2009, he joined the Department ofElectronic Engineering, Macquarie Uni-versity, Sydney, Australia, where he isnow a Lecturer. Dr Gengfa Fang’s re-

search interests include MAC protocols design and optimiza-tion for Wireless Body Area Networks, Wireless BroadbandNetworks and Cognitive Wireless Networks; Queue theory andoptimization, packet scheduling and cross-layer design.

Eryk Dutkiewicz received his B.E. de-gree in Electrical and Electronic En-gineering from the University of Ade-laide in 1988, his M.S. degree in Ap-plied Mathematics from the Universityof Adelaide in 1992 and his PhD inTelecommunications from the Univer-sity of Wollongong in 1996. His industryexperience includes management of theWireless Research Laboratory at Mo-torola in early 2000’s. He is currently

a Professor of Wireless Communications in the Departmentof Electronic Engineering, Macquarie University, Sydney, Aus-tralia. He has held visiting professorial appointments at sev-eral institutions including the Chinese Academy of Sciencesand the Shanghai JiaoTong University. His current researchinterests cover medical body area networks and cognitive radionetworks.

interference study of medical body area network for ... · interference study of medical body area...

Documents