On the Implant Communication and MAC Protocols for a WBAN

Abstract

Recent advances in micro-electro-mechanical systems (MEMS), wireless communication, low-power intelligent sensors, and semiconductor technologies have allowed the realization of a wireless body area network (WBAN). A WBAN provides unobtrusive health monitoring for a long period of time with real-time updates to the physician. It is widely used for ubiquitous healthcare, entertainment, and military applications. The implantable and wearable medical devices have several critical requirements such as power consumption, data rate, size, and low-power medium access control (MAC) protocols. This article consists of two parts: body implant communication, which is concerned with the communication to and from a human body using RF technology, and WBAN MAC protocols, which presents several low-power MAC protocols for a WBAN with useful guidelines. In body implant communication, the in-body radio frequency (RF) performance is affected considerably by the implant's depth inside the human body as well as by the muscle and fat. We observe best performance at a depth of 3cm and not close to the human skin. Furthermore, the study of low-power MAC protocols highlights the most important aspects of developing a single, a low-power, and a reliable MAC protocol for a WBAN.

Keyword. :

In-body, on-body, RF communication, Implant, WBAN

1. Introduction

Cardiovascular diseases are the foremost cause of deaths in the United States and Europe since 1900. More than ten million people are affected in Europe, one million in the US, and twenty two million people in the world [1]. The number is projected to be triple by 2020, resulting in an expenditure of around 20% of the gross domestic product (GDP). The ratio is 17% in South Korea and 39% in the UK [2]. The healthcare expenditure in the US is expected to be increased from $2.9 trillion in 2009 to $4 trillion US dollars in 2015 [3]. The impending health crisis attracts researchers, industrialists, and economists towards optimal and quick health solutions. The non-intrusive and ambulatory health monitoring of patient's vital signs with real time updates of medical records via internet provide economical solutions to the health care systems.

A wireless body area network (WBAN) is becoming increasingly important for healthcare systems, sporting activities, and members of emergency as well as military services. WBAN is an integration of in-body (implants) and on-body (wearable) sensors that allow inexpensive, unobtrusive, and long-term health monitoring of a patient during normal daily activities for prolonged periods of time. In-body radio frequency (RF) communications have the potential to dramatically change the future of healthcare. For example, they allow an implanted pacemaker to regularly transmit performance data and the patient's health status to the physician. However, the human body poses many wireless transmission challenges. This is partially conductive and consists of materials having different dielectric constants and characteristics impedance. The interface of muscles and fats may reflect the RF wave rather than transmitting it. The key elements of an RF-linked implant are the in-body antenna and the communication link performance. Also, in the case of many implants and wearable sensors, a low-power MAC protocol is required to accommodate the heterogeneous traffic in a power-efficient manner. This article is divided into two parts: body implant communication and WBAN MAC protocols. In the body implant communication part, we look at the RF communication link performance at various depths inside a human (artificial) body. In the MAC part, we review the existing low-power MAC protocols and discuss their pros and cons in the context of a WBAN. We further provide alternative MAC solutions for in-body and on-body communication systems.

The rest of the article is divided into three sections. In section 2, we present a discussion on body implant communication including in-body electromagnetic induction, RF communication, antenna design, and the communication link performance. Section 3 discusses several low-power MAC protocols and realizes a need for a new, a low-power, and a reliable MAC protocol for a WBAN. The final section concludes our work.

2. Body Implant Communication

There are several ways to communicate with an implant that includes the use of electromagnetic induction and RF technology. Both are wireless and their use depends on the application requirements. Further, the key elements of an RF-linked implant are the in-body antenna and the communication link performance. The following part discusses in-body electromagnetic induction, RF communication, antenna design, and the communication link performance.

2.1. In-body Electromagnetic Induction

Several applications still use electromagnetic coupling to provide a communication link to an implant device. In this scheme, an external coil is held very close to the body that couples to a coil implanted just below the skin surface. The implant is powered by the coupled magnetic field and requires no battery for communication. Data is transferred from the implant by altering the impedance of the implanted loop that is detected by the external coil and electronics. This type of communication is commonly used to identify animals that have been injected with an electronic tag. Electromagnetic induction is used when continuous, long-term communication is required. The base band for electromagnetic communication is typically 13.56 MHz or 28 MHz, with other frequencies also available. The choice of a particular band is subject to regulation for maximum specific absorption rate (SAR). The inductive coupling achieves best power transfer efficiency when uses large transmit and receive coils. It, however, becomes less efficient when the space is an issue of the device is implanted deep inside the human body. Furthermore, inductive coupling technique does not support a very high data rate and cannot initiate a communication session from inside of the body.

2.2. In-body RF Communication

Compared with the electromagnetic induction, RF communication dramatically increases bandwidth and supports a two-way data communication. The band designated for the in-body RF communication is medical implant communication service (MICS) band and is around 403 to 405 MHz. This band has a power limit of 25 µW in the air and is usually split into ten channels of 300 kHz bandwidth each.

The human body is a medium that poses numerous wireless transmission challenges. It consists of various components that are not predictable and will change as the patient ages, gains or losses weight, or even changes posture. Values of dielectric constant (εr), conductivity (σ) and characteristic impedance (Zo) for some body tissue are given in table 1 [4]. This demonstrates that these two tissue types are very different. Also, the dielectric constant affects the wavelength of a signal. At 403 MHz, the wavelength in the air is 744mm, but in muscle with εr = 50 the wavelength reduces to 105mm, which helps in designing implanted antennas.

2.3. In-body Antenna Design

A modern in-body antenna should be tuneable by using an intelligent transceiver and software routine. This enables the antenna coupling circuit to be optimised. Due to the frequency, and available volume, a non-resonant antenna is commonly used. It has a lower gain than a resonant antenna. This makes design of the antenna coupling circuit very important.

Antenna options are dictated by the location of the implant. A patch antenna can be used when the implant is flat. Patch antennas are comprised of a flat insulating substrate coated on both sides with a conductor. The substrate is a body compatible material with a platinum or a platinum/iridium conductor. The upper surface is the active face and is connected to the transceiver. The connection to the transceiver needs to pass through the case where the hermetic seal is maintained, requiring a feed-through. The feed-through must have no filter capacitors present; these are common on other devices. An implanted patch antenna is electrically larger than its physical size because it is immersed in a high (εr) medium. It can be much larger electrically if the substrate is of higher (εr), such as titania or zirconia.

A loop antenna can also be attached to the implant. This antenna operates mostly by the magnetic field, whereas the patch operates mostly by the electric field. The loop antenna delivers performance comparable to that of a dipole, but with a considerably smaller size. In addition, the magnetic permeability of muscle or fat is very similar to that of an air, unlike the dielectric constant that varies considerably. This property enables an antenna to be built and used with much less need for retuning. A loop antenna can be mounted on the case in a biocompatible structure.

2.4. In-body Link Performance

The demonstration system consists of a base-station, an implant, antennas, and a controlling laptop. The base-station contains a printed circuit board (PCB) with a wakeup RF circuit, a Zarlink ZL70101 IC, and a micro-controller. It sends a wakeup signal on industrial, scientific, and medical (ISM) 2.4 GHz band to power up the implant to communicate. It also supports communication within the MICS band. The implant contains a Zarlink ZL70101 IC, a micro-controller, and a battery. The power limits of the wakeup signal for ISM and MICS bands transmitters are 100mW and 25 µW respectively.

Experiments that measure the performance of an implant inside a living body are difficult to arrange. The alternative is to use 3D simulation software or a body phantom defined in [5]. The use of 3D simulation software is time consuming and hence practically not valuable. Therefore, measurements are generally performed using the body phantom and immersing a battery-powered implant into it [6]. Since no additional cables are attached to the test implant, the interference errors in the measurements are minimal. The body phantom is filled with a liquid that mimics the electrical properties of the human body tissues. The test environment is an anechoic chamber that includes a screened room. The interior walls of the room have sound-absorbent cones to minimize any reflections from walls or the floor that could distort the results. In real life, however, the results will be affected by the reflections from walls, desks, and other equipment and hardware. The body phantom is mounted on a wooden stand (non-conductive). The distance from the body phantom to the base-station is 3m. The MICS base-station dipole antenna is mounted on a stand. ' 1(a) shows the anechoic chamber with a body phantom (on the wooden stand), a log periodic test antenna (foreground), and a base-station dipole (right). The log periodic antenna is used to calculate the power radiated from the body phantom. A depth is defined as the horizontal distance between the outer skin of the phantom and the test implant. Vertical polarization of the implant is the case when the long side of the box and the patch antenna is vertical.

The link performance is measured once the communication link is established. The measurements include the effective radiated power (ERP) from the implant, the received signal at the implant from the base-station, and the link quality. Measurements are made over a set distance with all the combinations of implant and test antenna polarisations, i.e., vertical-vertical (V-V), horizontal-vertical (H-V), vertical-horizontal (V-H), and horizontal-horizontal (H-H) polarisations. Typical results are shown in ' 1(b) where the ERP is calculated from the received signal power and the antenna characteristics. The measurement of the signal levels is done with the log periodic antenna and the spectrum analyzer. It can be seen in the ' that there is a significant difference in signal levels with polarisation combinations and depths. For a V-V polarisation, the ERP increases from a 1cm depth to a maximum between 2 and 7 cm, and then it decreases. The gradual increase is due to the simulated body acting as a parasitic antenna. The ' also shows how the signal level is affected by the depth with different polarisation. Such a test needs to be done with the antenna that is to be used in the final product.

To measure the received signal at the implant, the Zarlink ZL70101 has an inbuilt receive signal strength indication (RSSI) function that gives a measure of the signal level detected. RSSI is a relative measurement with no calibration. The implant receives and measures a continuous wave signal transmitted by the base-station. In this case, the implant and the base-station antennas are vertically polarised. ' 1(c) shows an increase in the signal level at a depth between 3 and 4cm for a 15dec power. The power settings refer to the base-station and are con'd to set the ERP to 25 µW.

Signal levels are not valuable unless they are related to data transmission. One way to maintain the link quality is to measure the number of times the error correction is invoked during the transmission of 100 blocks of data. Two types of error correction codes, i.e., error correction code (ECC) and cyclic redundancy code (CRC) are invoked to maintain data integrity and reliability. The fewer ECC and CRC invocations result in better link quality. In ' 1(d), the error correction is lowest at a depth between 3 and 5 cm. A sample of ECC data collected at a 3cm implant depth is given in Table 2. The Count indicates the number of data blocks, the Time (ms) indicates the block transmission time, and the ECC indicates the number of times it is invoked. During the transmission of 100 blocks of data at a 3cm depth, the ECC is invoked 368 times, which is further equivalent to an average 3.68 times (as given in ' 1(d)).

2.5. Discussion

The ERP, RSSI, as well as the ECC and CRC plots show that the implant demonstrates the best performance at a depth between 3 and 5 cm. The depth and position of an implant is not chosen for engineering performance but for the best clinical reasons. The implant designer must be aware of the possible losses through the human body. The attenuation and the parasitic antenna effects vary from patient to patient, with the position of the implant and with the time as the patient gains, or looses weight. Therefore, these factors need to be built into the link budget.

3. WBAN MAC Protocols

Some of the common objectives in a WBAN are to achieve maximum throughput, minimum delay, and to maximize the network lifetime by controlling the main sources of energy waste, i.e., collision, idle listening, overhearing, and control packet overhead. A collision occurs when more than one packet transmits data at the same time. The collided packets have to be retransmitted, which consumes extra energy. The second source of energy waste is idle listening, meaning that a node listens to an idle channel to receive data. The third source is overhearing, i.e., to receive packets that are destined to other nodes. The last source is control packet overhead, meaning that control information area added to the payload. Minimal number of control packets should be used for data transmission.

Generally MAC protocols are grouped into contention-based and schedule-based MAC protocols. In contention-based MAC protocols such as carrier sense multiple access/collision avoidance (CSMA/CA) protocols, nodes contend for the channel to transmit data. If the channel is busy, the node defers its transmission until it becomes idle. These protocols are scalable with no strict time synchronization constraint. However, they incur significant protocol overhead. In schedule-based protocols such as time division multiple access (TDMA) protocols, the channel is divided into time slots of fixed or variable duration. These slots are assigned to nodes and each node transmits during its slot period. These protocols are energy conserving protocols. Since the duty cycle of radio is reduced, there is no contention, idle listening and overhearing problems. But these protocols require frequent synchronization. Table 3 compares CSMA/CA and TDMA protocols.

3.1. WBAN MAC Requirements

The most important attribute of a good MAC protocol for a WBAN is energy efficiency. In some applications, the device should support a battery life of months or years without interventions, while others may require a battery life of tens of hours due to the nature of the applications. For example, cardiac defibrillators and pacemakers should have a lifetime of more than 5 years, while swallowable camera pills have a lifetime of 12 hours. Power-efficient and flexible duty cycling techniques are required to minimize the idle listening, overhearing, packet collisions and control packet overhead. Furthermore, low duty cycle nodes should not receive frequent synchronization and control information (beacon frames) if they have no data to send or receive.

The WBAN MAC should also support simultaneous operation on in-body (MICS) and on-body channels (ISM or UWB) at the same time. In other words, it should support multiple physical layer (Multi-PHYs) communication or MAC transparency. Other important factors are scalability and adaptability to changes in the network, delay, throughput, and bandwidth utilization. Changes in the network topology, the position of the human body, and the node density should be handled rapidly and successfully. The MAC protocol for a WBAN should consider the electrical properties of the human body and the diverse traffic nature of in-body and on-body nodes. For example, the data rate of in-body nodes varies, ranging from few kbps in pacemaker to several Mbps in capsular endoscope.

In the following sections, we discuss proposed MAC protocols for a WBAN with useful guidelines. We also present a case study of IEEE 802.15.4, PB-TDMA, and S-MAC protocols for a WBAN using NS2 simulator.

3.2. Proposed MAC Protocols for a WBAN

In this section, we study proposed MAC protocols for a WBAN followed by useful suggestions/comments. Many of the proposed MAC protocols are the extension of existing MAC protocols originally proposed for wireless sensor networks (WSNs).

3.2.1. IEEE 802.15.4

IEEE 802.15.4 has remained the main focus of many researchers during the past few years. Some of the main reasons of selecting IEEE 802.15.4 for a WBAN were low-power communication and support of low data rate WBAN applications. Nicolas et.al investigated the performance of a non-beacon IEEE 802.15.4 in [7], where low upload/download rates (mostly per hour) are considered. They concluded that the non-beacon IEEE 802.15.4 results in 10 to 15 years sensor lifetime for low data rate and asymmetric WBAN traffic. However, their work considers data transmission on the basis of periodic intervals which is not a perfect scenario in a real WBAN. Furthermore, the data rate of in-body and on-body nodes are not always low, i.e., it ranges from 10 Kbps to 10 Mbps, and hence reduces the lifetime of the sensor nodes. Li et.al studied the behavior of slotted and unslotted CSMA/CA mechanisms and concluded that the unslotted mechanism performs better than the slotted one in terms of throughput and latency but with high cost of power consumption [8].

Intel Corporation conducted a series of experiments to analyze the performance of IEEE 802.15.4 for a WBAN [9]. They deployed a number of Intel Mote 2 [10] nodes on chest, waist, and the right ankle. Table 4 shows the throughput at a 0dBm transmit power when a person is standing and sitting on a chair. The connection between ankle and waist cannot be established, even for a short distance of 1.5m. All other connections show favourable performance.

Dave et al. studied the energy efficiency and QoS performance of IEEE 802.15.4 and IEEE 802.11e [11] MAC protocols under two generic applications: a wave-form real time stream and a real-time parameter measurement stream [12]. Table 5 shows the throughput and the Power (in mW) for both applications. The AC_BE and AC_VO represent the access categories voice and best-effort in the IEEE 802.11e.

Since the IEEE 802.15.4 operates in the 2.4 GHz unlicensed band, the possibilities of interference from other devices such as IEEE 802.11 and microwave are inevitable. A series of experiments to evaluate the impact of IEEE 802.11 and microwave ovens on the IEEE 802.15.4 transmission are carried out in [13]. The authors considered XBee 802.15.4 development kit that has two XBee modules. Table 6 shows the affects of microwave oven on the XBee remote module. When the microwave oven is ON, the packet success rate and the standard deviation is degraded to 96.85% and 3.22% respectively. However, there is no loss when the XBee modules are taken 2 meters away from the microwave oven.

3.2.2. Heartbeat Driven MAC Protocol (H-MAC)

A Heartbeat Driven MAC protocol (H-MAC) [14] is a TDMA-based protocol originally proposed for a star topology WBAN. The energy efficiency is improved by exploiting heartbeat rhythm information in order to synchronize the nodes. The nodes do not need to receive periodic information to perform synchronization. The heartbeat rhythm can be extracted from the sensory data and hence all the rhythms represented by peak sequences are naturally synchronized. The H-MAC protocol assigns dedicated time slots to each node to guarantee collision-free transmission. In addition, this protocol is supported by an active synchronization recovery scheme where two resynchronization schemes are implemented.

Although H-MAC protocol reduces the extra energy cost required for synchronization, it does not support sporadic events. Since the TDMA slots are dedicated and not traffic adaptive, H-MAC protocol encounters low spectral/bandwidth efficiency in case of a low traffic. For example, a blood pressure node may not need a dedicated time slot while an endoscope pill may require a number of dedicated time slots when deployed in a WBAN. But the slots should be released when the endoscope pill is expelled. The heartbeat rhythm information varies depending on the patient condition. It may not reveal valid information for synchronization all the time. One of the solutions is to assign the time slots based on the node's traffic information and to receive synchronization packets when required, i.e., when a node has data to transmit/receive.

3.2.3. Reservation-based Dynamic TDMA Protocol (DTDMA)

A Reservation-based Dynamic TDMA Protocol (DTDMA) [15] is originally proposed for a normal (periodic) WBAN traffic where slots are allocated to the nodes which have buffered packets and are released to other nodes when the data transmission/reception is completed. The channel is bounded by superframe structures. Each superframe consists of a beacon - used to carry control information including slot allocation information, a CFP period - a configurable period used for data transmission, a CAP period - a fixed period used for short command packets using slotted aloha protocol, and a configurable inactive period - used to save energy. Unlike a beacon-enabled IEEE 802.15.4 superframe structure where the CAP duration is followed by CFP duration, in DTDMA protocol the CFP duration is followed by CAP duration in order to enable the nodes to send CFP traffic earlier than CAP traffic. In addition, the duration of inactive period is configurable based on the CFP slot duration. If there is no CFP traffic, the inactive period will be increased. The DTDMA superframe structure is given in ' 2(a).

It has been shown that for a normal (periodic) traffic, the DTDMA protocol provides more dependability in terms of low packet dropping rate and low energy consumption when compared with IEEE 802.15.4. However, it does not support emergency and on-demand traffic. Although the slot allocation based on the traffic information is a good approach, the DTDMA protocol has several limitations when considered for the MICS band. The MICS band has ten channels where each channel has 300 Kbps bandwidth. The DTDMA protocol is valid only for one channel and cannot operate on ten channels simultaneously. In addition, the DTDMA protocol does not support the channel allocation mechanism in the MICS band. This protocol can be further investigated for the MICS band by integrating the channel information in the beacon frame. The new concept may be called Frequency-based DTDMA (F-DTDMA), i.e., the coordinator first selects one of the channels in the MICS band and then divides the selected channel in TDMA superframe (s) according to the DTDMA protocol. However the FCC has imposed several restrictions on the channel selection/allocation mechanism in the MICS band, which further creates problems for the MAC designers.

3.2.4. BodyMAC Protocol

A BodyMAC protocol is a TDMA-based protocol where the channel is bounded by TDMA superframe structures with downlink and uplink subframes as given in ' 2(b) [16]. The downlink frame is used to accommodate the on-demand traffic and the uplink frame is used to accommodate the normal traffic. There is no proper mechanism to handle the emergency traffic. The uplink frame is further divided into CAP and CFP periods. The CAP period is used to transmit small size MAC packets. The CFP period is used to transmit the normal data in a TDMA slot. The duration of the downlink and uplink superframes are defined by the coordinator.

The advantage of the BodyMAC protocol is that it accommodates the on-demand traffic using the downlink subframe. However, in case of low-power implants (which should not receive beacons periodically), the coordinator has to wake up the implant first and then send synchronization packets. After synchronization, the coordinator can request/send data in the downlink subframe. The wake up procedure for low-power implants is not defined in the BodyMAC protocol. One of the solutions is to use a wakeup radio in order to wake up low-power implants before using the downlink subframe. In addition the wakeup packets can be used to carry control information such as channel (MICS band) and slot allocation information from the coordinator to the nodes. Finally, the BodyMAC protocol uses the CSMA/CA protocol in the CAP period which is not reliable for a WBAN. This should be replaced by slotted-ALOHA as done in DTDMA.

Further details on low-power MAC protocols (originally proposed for WSNs) for a WBAN are given in Appendix I.

3.3. Case Study: IEEE 802.15.4, PB-TDMA, and SMAC Protocols for a WBAN

In this section, we investigate the performance of a beacon-enabled IEEE 802.15.4, preamble-based TDMA [17], and SMAC protocols for an on-body communication system. Our analysis is verified by extensive simulations using NS-2. The wireless physical parameters are considered according to a low-power Nordic nRF2401 transceiver (Chipcon CC2420 radio [18] is considered in case of IEEE 802.15.4) [19]. This radio transceiver operates in the 2.4-2.5 GHz band with an optimum transmission power of -5dBm. We use the shadowing propagation model throughout the simulations. We consider a total of 7 nodes firmly placed on a human body. The nodes are connected to the coordinator in a star topology. The distribution of the nodes and the coordinator is given in ' 3(a). The initial node's energy is 5 Joules. The packet size is 120 bytes. The average data transmission rate of ECG, EEG, and EMG is 10, 70, and 100 kbps. The transport agent is a user datagram protocol (UDP). Since the traffic is an uplink traffic, the buffer size at the coordinator is considered unlimited. In a real WBAN, the buffer size should be configurable based on the application requirements. For energy calculation, we use the existing energy model defined in NS-2. The simulation area is 3x3 meter and each node generates constant bit rate (CBR) traffic. The CBR traffic is an ideal model for some of the medical applications, where the nodes send data based on pre-defined traffic patterns. However, most of the nodes in a WBAN have heterogeneous traffic characteristics and they generate periodic and aperiodic traffic. In this case, they will have many traffic models operating at the same time, ranging from CBR to variable bit rate (VBR).

' 3(b) shows the throughput of the IEEE 802.15.4, PB-TDMA, and S-MAC protocols. The performance of the IEEE 802.15.4, when con'd in a beacon-enabled mode, outperforms PB-TDMA and S-MAC protocols. The efficiency of a MAC protocol depends on the traffic pattern. In this case, S-MAC protocol results poor performance because the traffic scenario that we generated is not an ideal scenario for the S-MAC. ' 3(c) shows the residual energy at various nodes during simulation time. When nodes finish their transmission, they go into sleep mode, as indicated by the horizontal line. The coordinator has a considerable energy loss because it always listens to the other nodes. However, the energy consumption of the coordinator is not a critical issue in a WBAN. We further analyze the residual energy at the ECG node for different transmission powers. There is a minor change in energy loss for three different transmission powers as given in ' 3(d). This concludes that reducing the transmission power only does not save energy unless supported by an efficient power management scheme.

The IEEE 802.15.4 can be considered for certain on-body medical applications, but it does not achieve the level of power required for in-body nodes. It is not sufficient for high data rate medical and non-medical applications due to its limitations to 250 kbps. Furthermore, it uses slotted or unslotted CSMA/CA where the nodes are required to sense the channel before transmission. However, the channel sensing is not guaranteed in MICS band because the path loss inside the human body due to tissue heating is much higher than in free space. Bin et.al studied the clear channel assessment (CCA) range of in-body nodes which is only 0.5 meters [20]. This unreliability in CCA indicates that CSMA/CA is not an ideal technique for the in-body communication system. An alternative approach is to use a TDMA-based protocol that contains a beacon, a configurable contention access period (CCAP), and a contention free period (CFP) [21]. Unlike the IEEE 802.15.4, this protocol is required to use a slotted-ALOHA protocol in the CCAP instead of CSMA/CA. The CCAP period should contain few slots (3 or 4) of equal duration and can be used for short data transmission and a guaranteed time slot (GTS) allocation. To enable a logical connection between the in-body and the on-body communication systems, a method called bridging function can be used as discussed in [21]. The bridging function can integrate in-body and on-body nodes into a WBAN, thus satisfying the MAC transparency requirement. Further details about bridging function are given in [22].

3.4. Discussion

Since the CSMA/CA is not suitable due to unreliable CCA and heavy collision problems, it can be seen that the most reliable power-efficient protocol is a TDMA-based protocol. Many protocols have been proposed for a WBAN and most of them are based on a TDMA-based mechanism. However, all of them have pros and cons for a real WBAN system that should operate on Multi-PHYs (MICS, ISM, and UWB) simultaneously. The MAC transparency has been a hot topic for the MAC designers since different bands have different characteristics in terms of data rate, number of channels in a particular frequency band, and data prioritization. A good MAC protocol should enable reliable operation on MICS, ISM, and UWB etc bands simultaneously. The main problems are related to MICS band due to FCC restrictions [23]. According to FCC,

“Within 5 seconds prior to initiating a communications session, circuitry associated with a medical implant programmer/control transmitter must monitor the channel or channels the MICS system devices intend to occupy for a minimum of 10 milliseconds per channel.”

In other words, the coordinator must perform Listen-before-talking (LBT) criteria prior to a MICS communication sessions. The implants are not allowed to select a channel and this further creates problem for emergency traffic. Since the channels are solely assigned by the coordinator, the emergency implants have to wait until the channels are assigned to them. This restriction prevents MAC designers to develop a reliable mechanism for emergency traffic. One of the solutions is to use a wakeup radio or a control channel dedicated to emergency traffic only. However, the FCC does not allow the dedication of a channel for a wakeup radio or for a control channel. Also, the use of a separate antenna for a wakeup radio is not suitable for the implants.

In a nutshell, the unique characteristics of a WBAN require the design and implementation of a novel, low-power, and heterogeneous MAC protocol that should satisfy the traffic heterogeneity and correlation, MAC transparency, and reliability requirements.

4. Conclusions

This article investigated the performance of an RF-linked implant at various depths inside a human (artificial) body. A potential key element in an efficient RF communication is the placement of the implanted device. In general, a medical examiner places the implant where it is clinically effective with little concern for RF propagation. In practice, it is required that the implant should operate effectively at various in-body depths and through the layers of fat, muscle, and skin. Based on our measurement results, we observed the best performance at a 3 cm depth from the body surface. However, a data link can be maintained at a depth of over 15 cm. We further studied several low-power MAC protocols for a WBAN and addressed potential issues and challenges in the MAC protocol design for in-body and on-body communication systems. Most of the existing low-power MAC protocols accommodate certain WBAN applications but are not sufficient to satisfy the stringent low-power requirements of in-body nodes. Therefore, further study is required to propose a new low-power and reliable MAC protocol for in-body and on-body communication systems. The new protocol should satisfy the traffic heterogeneity and correlation, MAC transparency, low-power communication, and reliability requirements.

References

[1] Cleland JG, Swedberg K., Follath F., "A Survey of the Quality of Care Among Patients with Heart Failure in Europe. Part 1: Patient Characteristics and Diagnosis" The Euro Heart Failure Survey Programme, pp 24:442-463, Eur. Heart. J. 2003.

[2] http://www.who.int/whosis/mort/profiles/mortwprokorrepofkorea.pdf Data visited: 3/12/2008

[3] Borger, C., et al., Health Spending Projections Through 2015: Changes on the Horizon. In Health Affairs Web Exclusive W61 February 22, 2006.

[4] Yang G-Z, “Body Sensor Networks, Springer,” pp 117-143, 2006.

[5] Electromagnetic Compatibility and Radio Spectrum Matters (ERM); Electromagnetic Compatibility (EMC); Standard for Radio Equipment and Services; Part 27; Specific Conditions For Ultra Low Power Medical Implants (ULP - AMI) and Related Peripheral Devices(ULP-AMI-P)”, ETSI EN301 489-27, V1.1.1, ETSI, March 2003.

[6] Higgins, H., "Body Implant Communication¿ Making it Possible," The Second European Conference on Antennas and Propagation, pp.1-4, 11-16 Nov. 2007

[7] Nicholas F. Timmons, William G. Scanlon, ``Analysis of the performance of IEEE 802.15.4 for medical sensor body area networking'', IEEE SECON (2004), 2004.

[8] Changle Li, Huan-Bang Li, Kohno, R., "Performance Evaluation of IEEE 802.15.4 for Wireless Body Area Network (WBAN)," Communications Workshops, 2009. ICC Workshops 2009. IEEE International Conference on , vol., no., pp.1-5, 14-18 June 2009.

[9] Shah, R.C., Yarvis, M., "Characteristics of on-body 802.15.4 networks," Wireless Mesh Networks, 2006. WiMesh 2006. 2nd IEEE Workshop on , vol., no., pp.138-139, 25-28 Sept. 2006.

[10]. R. Adler et. al., “Intel Mote 2: An advanced platform for demanding sensor network applications”, Demo session, Sensys 2005.

[11] IEEE 802.11e Std., “Amendment to Part 11: “Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications,”: “Medium Access Control Quality of Services Enhancements,” November 2005.
[12] Dave Cavalcanti, Ruediger Schmitt and Amjad Soomro, “Performance Analysis of 802.15.4 and 802.11e for Body Sensor Network Applications,” 4th International Workshop on Wearable and Implantable Body Sensor Networks, BSN 2007.

[13] Chao Chen, Carlos Pomalaza-Ráez ,Monitoring Human Movements at Home Using Wearable Wireless Sensors, ISMICT 2009, Montreal, 24-27 February, 2009.

[14] Huaming Li H Jindong Tan H , Heartbeat Driven Medium Access Control for Body Sensor Networks, Proceedings of the 1st ACM SIGMOBILE international workshop on Systems and networking support for healthcare and assisted living environments, pg 25-30, 2007.

[15] Changle LI, Huan-Bang LI and Ryuji KOHNO, “Reservation-Based Dynamic TDMA Protocol for Medical Body Area Networks”, IEICE Trans. Commun., Vol.E92.B, No.2, pp.387-395, 2009.

[16] Gengfa Fang, Eryk Dutkiewicz, "BodyMAC: Energy Efficient TDMA-based MAC Protocol for Wireless Body Area Networks" in IEEE ISCIT 2009, Incheon, Korea, September 28-30, 2009.

[17] Sana Ullah, Riazul Islam, Ahasanun Nessa, Yingji Zhong, Kwak Kyung Sup, Performance Analysis of a Preamble Based TDMA Protocol for Wireless Body Area Network, Journal of Communications Software and Systems, Vol 4, No 3,pg 222-226, November 2008.

[18] http://www.moteiv.com. Date visited: July 2008.

[19] http://www.sparkfun.com/datasheets/RF/nRF2401rev1_1.pdf Date visited: 12/11/2008

[20] Bin Zhen, Huan-Bang Li, and Ryuji Kohno, IEEE body area networks and medical implant communications, Proceedings of the ICST 3rd international conference on Body area networks, Tempe, Arizona, 2008.

[21] 802.15-09-0366-00-0006, A Traffic-based Secure MAC Protocol for WBAN with Bridging Function, IEEE Standards Meeting (TG6), Montreal, Canada, May 2009.

[22] Sana Ullah, Xizhi An, Kyung Sup Kwak, "Towards Power-Efficient MAC Protocol for In-body and On-body sensor Networks", KES AMSTA 09, LNAI 5559, pp.335-345, Uppsala, June 2009

[23] July 2009, http://www.fcc.gov/Bureaus/Wireless/Orders/1999/fcc99363.txt

Appendix I

We study the following protocols and analyzed their pros and cons for a WBAN. They are not explained here due space limitation problems

Protocols

Channels

Organization and Basic Operation

Advantages and Disadvantages

Adaptability to WBANs/Comments

WiseMAC

1

Organized randomly and operation is based on listening

Scalable and adaptive to traffic load, Support mobility, low and high power consumption in low and high traffic conditions, and low delay

Good for high traffic applications, not suitable for low duty cycle in-body/on-body nodes

BMAC

1

Organized in slots and operation is based on schedules

Flexible, high throughput, tolerable latency, and low power consumption

Good for high traffic applications

STEM

2

Organized randomly having two channels (control + data channel) and operation is based on wakeup schedules

Suitable for events based applications

Good for periodic traffic especially for low traffic applications. Suitable to handle sporadic events due to a separate control channel. But hard to handle sporadic events when the traffic load is high

SMAC

1

Organized in slots and operation is based on schedules

High transmission latency, loosely synchronized, low throughput

Good for high traffic applications. Suitable for applications where throughput is not a primary concern such as in-body medical applications

TMAC

1

Organized in slots and operation is based on schedules

Queued packets are sent in a burst thus achieve better delay performance, loosely synchronized

Good for high traffic applications.

Early sleep problems allow the nodes to loose synchronization

PMAC

1

Organized in hybrid mode and operation is based on listening

Adaptation to changes might be slow, loosely synchronized, high throughput under heavy traffic

Good for delay-sensitive applications

DMAC

1

Organized in slots and operation is based on schedules

better delay performance due to Sleep schedules, loosely synchronized, optimized for data forwarding sink

On-body nodes can be prioritized according to their application requirements and a data tree can be built, where the WBAN coordinator will be a cluster node

FLAMA

1

Organized in frames and operation is based on schedules

Better end-to-end reliability and energy saving, smaller delays, improved energy saving, high reliability

Good for low power applications. Adaptable to high traffic applications.

LEACH

1

Organized in clusters and operations is based on TDMA scheme

Distributed protocol, no global knowledge required, extra overhead for dynamic clustering

TDMA schedules should be created by the WBAN coordinator. Cluster head should not change (depending on minimum communication energy) as in the traditional LEACH

HEED

1

Organized in clusters and operations is based on TDMA scheme

Good for energy efficiency, scalability, prolonged network lifetime, load balancing

The WBAN coordinator acts as a cluster head. Unlike traditional HEED, the WBAN network size is often defined (by the physician)

All of the above protocols are designed for a single channel. However, a MAC protocol for a WBAN should operate on multiple channels simultaneously. In the above table, we looked at each protocol and gave miscellaneous comments for a WBAN. None of the above protocols satisfy the entire WBAN MAC requirements. However, this study can allow the MAC designer to choose the best and ignore the worst from the above methods and to develop a novel protocol that contains the best of all protocols.

Frequency

Muscle

Fat

()

(S.m-1)

()

(S.m-1)

100

66.2

0.73

31.6

12.7

0.07

92.4

400

58

0.82

43.7

11.6

0.08

108

900

56

0.97

48.2

11.3

0.11

111

Table 1. Body electrical properties [4]

Table 2. A sample of ECC data collected at a 3cm implant's depth.

Count

Time

(ms)

ECC

Count

Time

(ms)

ECC

Count

Time

(ms)

ECC

Count

Time

(ms)

ECC

Count

Time

(ms)

ECC

1

39.273

2

21

39.273

3

41

39.273

5

61

39.273

6

81

39.273

5

2

39.273

7

22

46.957

3

42

39.273

6

62

39.273

1

82

64.544

1

3

39.273

3

23

39.273

3

43

39.273

1

63

39.273

5

83

39.273

6

4

39.273

2

24

39.273

2

44

39.273

4

64

39.273

4

84

39.273

3

5

39.273

4

25

39.273

6

45

39.273

2

65

39.273

4

85

39.273

4

6

39.273

2

26

39.273

4

46

39.273

4

66

39.273

5

86

39.273

3

7

39.273

3

27

39.273

6

47

39.273

4

67

39.444

3

87

39.273

4

8

56.69

2

28

39.273

4

48

39.444

4

68

39.273

4

88

39.273

6

9

39.273

4

29

39.273

3

49

112.184

3

69

39.273

5

89

39.444

6

10

39.273

1

30

49.177

1

50

39.273

6

70

39.273

6

90

39.273

5

11

39.273

2

31

39.273

3

51

86.742

5

71

39.273

5

91

39.273

2

12

39.273

4

32

39.273

3

52

39.273

3

72

39.273

4

92

39.273

8

13

39.273

2

33

39.273

3

53

39.444

3

73

39.273

3

93

39.273

4

14

53.445

2

34

39.273

3

54

39.273

5

74

39.273

3

94

39.273

4

15

39.273

6

35

39.273

0

55

39.273

5

75

39.273

4

95

39.273

5

16

39.273

3

36

39.273

5

56

39.273

3

76

39.273

3

96

39.273

2

17

39.273

6

37

39.273

4

57

39.273

5

77

39.273

5

97

39.273

5

18

39.273

3

38

39.273

3

58

39.273

3

78

39.273

3

98

39.273

0

19

39.273

1

39

42.859

2

59

39.273

2

79

57.543

5

99

39.273

5

20

39.273

2

40

39.444

5

60

39.273

4

80

39.273

3

100

39.273

5

Maximum time (ms) = 112.184, Minimum time (ms) = 39.273, Average time (ms) = 41.448

Table 3. CSMA/CA VS. TDMA Protocols

Performance Metric

CSMA/CA

TDMA

Power consumption

High

Low

Traffic level

Low

High

Bandwidth utilisation

Low

Maximum

Scalability

Good

Poor

Effect of packet failure

Low

Latency

Synchronisation

Not Applicable

Required

Table 4. Throughput at a 0dBm Transmit Power

Throughput when a Person

is Standing

Throughput when a Person

is Sitting on an Office Chair

Source Nodes

Destination Nodes

Destination Nodes

Chest

Waist

Ankle

Chest

Waist

Ankle

Chest

-

99%

84%

-

99%

81%

Waist

100%

-

50%

99%

-

47%

Ankle

72%

76%

-

77%

27%

-

Table 5. Throughput and Power (in mW) of IEEE 802.15.4

and IEEE 802.11e under AC_BE and AC_VO

Sensor Nodes

IEEE 802.15.4

IEEE 802.11e

(AC_BE)

IEEE 802.11e

(AC_VO)

Throughput

Wave-form

100%

100%

100%

Parameter

99.77%

100%

100%

Power (mW)

Wave-form

1.82

4.01

3.57

Parameter

0.26

2.88

2.77

Table 6. Co-existence Test Results between

IEEE 802.15.4 and Microwave Oven

Microwave Status

Packet Success Rate

Mean

Std.

ON

96.85%

3.22%

OFF

100%

0%