Advertisement

Wireless Personal Communications

, Volume 100, Issue 4, pp 1263–1275 | Cite as

Channel Models for Body Surface Communications in Ultra Wideband-Based Wireless Body Area Networks

  • Deepak Kumar Rout
  • Susmita Das
Article
  • 148 Downloads

Abstract

Body area networks are being developed to serve a wide range of purposes ranging from providing health care to patients on the move to tracking patients and motion sensing for gaming controls. There has been significant and sizeable amount of research in the various areas and applications of body area networks. Ultra wideband which operates in the 3.1–10.6 GHz band is slowly being preferred for high data rate communication in body area networks. The development of suitable applications and techniques for communication depends significantly on the channel models. The wireless channel is a crucial parameter as it provides significant information about the propagation characteristics and losses involved in the transmission medium. The existing channel models proposed are mostly in the spectra involving the wideband 3.1–10.6 GHz bands or the 3.1–6 GHz bands. However, the IEEE 802.15.6 specifies operation in various sub-bands of 499.2 MHz width. And the channel characteristics are significantly different for wideband and narrowband channels. In this article, we propose empirical channel models for body surface communication in the various sub-bands specified by the IEEE 802.15.6. The body surface scenario is chosen as the combination of propagation through wireless media and losses due to absorption from body tissues make it challenging. The proposed path loss models are developed from more than 300,000 received power measurements collected over a span of hours.

Keywords

Body area network Ultra wideband Channel modelling Body surface communication 

1 Introduction

Body area network (BAN) is an amalgamation of sensor network, telemedicine and state of the art wireless communication techniques to provide remote health monitoring to patients on the move. The on body and implanted sensors collect health information in the form of electrocardiogram (ECG), electroencephalogram (EEG), blood pressure, body temperature, etc., and transmit the information wirelessly to a nearby medical monitoring center [1]. The algorithms implemented on the monitoring center server search for anomalies in the patient health information. The deviations from the normal health parameters indicate signs of diseases. Once an abnormality is detected the patient is located and health services are rushed. The BAN helps in the efficient utilization of medical personnel and resources that are scarce in several nations. Thus, researchers are working worldwide to develop techniques and make BANs implementable [2, 3, 4, 5, 6]. The primary areas of thrust have been biomedical engineering for detection of diseases, etc., wireless communication for the development of transceivers, and computer sciences for the development of various protocols.

The design and development of the wireless communication system of BAN is a major area of interest, as suggested by several researchers. The BAN system must be energy efficient, reliable, and of low complexity. The physical or radio layer of BAN requires high bandwidth, low transmit power yet reliable communication methodology [7, 8].

The ultra wideband (UWB) is an innovative form of communication and is physically different compared to the conventional systems which fits the BAN physical layer constraints discussed in several research papers [9, 10, 11, 12]. It contains an enormous 7.5 GHz of bandwidth in the 3.1–10.6 GHz band [4]. The large bandwidth helps the devices to use a pulse-based baseband transmission using a simple transceiver structure. Low power pulses are transmitted through the wireless medium which can be recovered using a correlation receiver of low complexity [4, 7]. However, before the UWB can be implemented in the physical layer of the BANs, it will require the measurement of radio propagation characteristics and the losses resulting from the phenomenon. A knowledge of the path loss will definitely help in the design of suitable power efficient transceivers [13, 14, 15, 16].

In conventional BAN systems, there are two distinct scenarios that exist for wireless communication. These are the body surface to body surface communication scenario and the body surface to external communication scenario [6, 17, 18]. Among the two environments the former is more treacherous in nature due to shadowing caused by absorption in the human body tissues, as reported by several eminent researchers [17, 19]. Thus, the approach in this article will be towards the body surface communication scenario.

The article has been divided into several sections for easy understanding and presentation. Initially, the article reviews the existing work in this area and discusses briefly the uniqueness of the proposed channel models. The section also presents the key motivation for the work. The succeeding section then provides a depiction of the measurement setup, equipment specifications and antenna characteristics which are vital in any measurement study. The channel measurement results are then investigated from the average and median path loss information.

2 Related Work

Channel modelling for body area networks is an active area of research and numerous researchers are working on investigating the channel channel characteristics in the UWB personal area networks [20, 21, 22, 23, 24]. In [25], Smith et al. probe the narrowband and wideband channels for BANs. Di Franco et al. [26] suggest that the BAN channels are quite diverse than existing channel models and the human body disturbs the signal propagation. An article by Molisch et al. [14] deliberates the channel models for UWB-based body area network. Although, the emphasis is particularly on personal area networks and the modelling is as per IEEE 802.15.3 standards, the findings are quite relevant with respect to the IEEE 802.15.6 standards. The article exposes the actual nature of the UWB-based BAN channels and the effects of human body on the path loss characteristics. Similarly, an article by Cotton et al. [18] offers an analysis of BAN channels for body centric communications. Pang et al. [15] suggest analytical channel models for wireless BANs. Similarly, an article by Taparugssanagorn et al. [16] present empirical channel models for ultra wideband-based BAN communications. The model is derived from real-time received power measurements in the 3–11 GHz band. The findings advocate that the complexity of human body tissues and the nature of propagation in this channel will be quite diverse from the conventional wireless channels. The most imperative concern with communication in and around the human body is the rapid degradation due to body tissues, which is advocated by Alomainy et al. [27] and Yazdandoost et al. [17] in their respective articles. An article by Zasowski et al. [10] put forward a noninvasive channel model in the 3–6 GHz band for UWB BANs in both anechoic chamber and office room scenarios.

The channel models proposed in the discussed research articles can be characterized into body surface to body surface and body surface to external scenarios. The articles presented prior, evaluate the existing channel models and discuss their pros and cons. Some of these articles also recommend improved channel models. While some of the channel models are focused on narrowband channels, reasonably a proportion are for wideband UWB channels. However, all of them lack in the aspect that they are not as per the IEEE 802.15.6 standards. These models are empirically formulated in the 3.1–6 GHz bands or the 3.1–10.6 GHz bands. However, the IEEE 802.15.6 recommends multiple sub-bands, each 499.2 MHz wide, but none of the proposed models deliberate this aspect of the IEEE standard. It is anticipated that the losses in these sub-bands will be quite diverse from the prevailing models due to the use of sub-bands, and hence, a measurement campaign becomes a way of studying the resulting deviations in the path loss.

Thus, in this article we have attempted to model the path loss in the body surface to body surface channels in the various frequency bands as suggested by the IEEE 802.15.6. The models are expressed as simple path loss equations and the median pathloss are provided for reference.

3 An Overview of the IEEE 802.15.6 Standards

The IEEE 802.15.6 is a personal area network standard optimized for body area networks. It is a standard issued by the Institution of Electrical and Electronics engineers, which designates the requirements and advocates the use of certain methods and protocols to ensure reliable and secure communication in and around the human body [3, 4]. The standard provides the specifications for the medium access control (MAC) and physical layer (PHY). It is highly bandwidth efficient and augmented for low power body nodes to serve a variety of medical and non-medical applications. The standard assigns several licensed and unlicensed bands for communication between the BAN nodes. It considers the preoccupation of the spectrum into account in the different regions of the world and suggests numerous sub-bands throughout the world. It specifies communication in three different PHYs namely, the narrowband PHY, the ultra wideband PHY, and the human body communication(HBC) PHY. The narrowband PHY makes use of conventional modulation techniques (phase shift keying, frequency shift keying, etc.) and operates in the 400, 600, 900 MHz and 2.4 GHz bands, while the HBC operates in the 50 MHz band using baseband methods that doesn’t require any modulation. The HBC communication also operates in two sub-bands at 16 and 27 MHz with each having a bandwidth of 4 MHz.

The ultra wideband physical layer, however, is of distinct interest due to the benefits discussed prior. It operates in several sub-bands within the 3.1–10.6 GHz spectra and the sub-bands are categorized into low band and high band, respectively. The sub-bands are separated by 499.2 MHz and identified with a channel number beginning from 0 and onwards. The channel numbers 0, 1 and 2 are with center frequencies of 3.4944, 3.9936 and 4.4928 GHz are designated as the low band while channels 3–10 are high band with center frequencies between 6.4896 and 9.984 GHz. In this paper, the three low band channels are investigated along with two other sub-bands with center frequencies 4.992 and 5.4912 GHz. These two bands are not mentioned in the IEEE 802.15.6 to avoid interference with existing systems. However, with suitable narrowband interference mitigation methods they can be used without much issues.

4 Channel Modelling Setup

The body surface to body surface (intra body) is the most prominent form of communication in the human body. In this setup, the transmitting and the receiving nodes are positioned on the human body surface or at most 2 cm above. The IEEE 802.15 task group 6 reports suggest that the body surface scenario presents the most adverse channel conditions due to absorption from body tissues and the shadowing resulting thereof [17]. Thus, it becomes important to understand the signal propagation in this scenario and study the path loss characteristics.

The experiment makes use of a setup similar to those used in the conventional path loss measurement campaigns. The measurement system was setup in a typical laboratory environment with wooden partitions surrounding the human body. The measurement setup consisted of a Vector signal generator (VSG) of Agilent make capable of generating signals up to 6 GHz, and a Spectrum Analyzer (SA) of Agilent make with up to 7 GHz measurement range. The transmitting and the receiving antennas were connected to the signal generator and the spectrum analyzer, respectively, effectively using low loss coaxial cables. The antennas were positioned on the subject’s body, the transmitter–receiver separation was adjustable to emulate the transmitter receiver separation, and the path loss were recorded. The subject chosen for the measurement campaign is an Asian male of age 30 years, weight 65 kg, and height 172 cm, the subject’s apparels included all cotton clothes. The antennas were never in direct contact with the human body and were always a 5 mm away from skin separated by the apparels. Figure 1 shows the measurement setup and highlights the placement positions of antennas. The gains of the transmitting and receiving antennas, and the loss due to the connecting cables were pre-calculated to derive the received power calculations, so as to evaluate the path loss. The signal was generated in MATLAB® environment and downloaded to the vector signal generator. The UWB signal generate by the signal generator was transmitted through the transmitting antenna over the human body surface. The signal after propagation through the periphery of the human body is received by the receiving antenna. The receiving antenna was linked to the spectrum analyzer so that the received power could be detected and measured.
Fig. 1

Measurement setup for body surface to body surface (intrabody) communication

4.1 Salient Features of Antenna Used for Measurement

The discussion on the channel propagation studies will be worthless without discussing the features of the antenna used for transmission and reception. In several research articles, the use of an miniature, omnidirectional antenna is proposed so that it doesn’t interfere with the normal lifestyle of the human being. Thus, the antenna used for the channel modeling setup is a miniscule microstrip antenna of dimensions 30 mm × 32 mm. The UWB antenna is designed to perform in the 2.6–10.3 GHz band and delivers gains in the range of 1.913–4.248 dB for the full range of frequencies. It offers gains of 2.549, 2.761, 2.949, 3.133 and 3.188 dB for the five frequency bands under consideration. Apart from it’s smaller size, the low cost, low complexity, omnidirectional pattern and high bandwidth are its advantages which makes it suitable for BAN applications. It can be easily embedded with any BAN transceiver for body surface communications. The antenna and its radiation pattern are displayed in Fig. 2.
Fig. 2

Antenna used for measurement and its radiation pattern

5 Channel Modelling Results

The transmitting and the receiving antenna were placed on the human body. The transmitter and receiver antenna displacement was initially fixed at 50 mm, and it was consistently varied to 1500 mm in the multiples of 50 mm. The reason for the choice of limiting the maximum distance to 1500 mm was considering the average height of the world population of about 1.7 m or 1700 mm. So, it justifies that the communication on the human body will involve a maximum transmitter–receiver separation of around 1500 mm at extreme. The subject was in standing posture away from the furniture and measurement equipment at a distance of approximately 1 m. The measurements were obtained for both line of sight (LOS) and non-line of sight (NLOS) setups. For each transmitter–receiver separation, a total of 1000 individual measurements were documented, which accounts for a total of 30,000 discrete measurements for each frequency band and scenario. Since there are a total of 5 distinct frequency bands and two different scenarios, it results in 3 × 105 discrete received power estimates. The collected path loss data were averaged and fitted by the curve fitting tool of MATLAB®. The parameters resulting from these received power estimates can be used to obtain a path loss model.

The detailed path loss, average path loss and fitted path loss are shown in Table 1 for the entire frequency bands and both line of sight and non-line of sight setups. In the 3.4944 GHz band, for the line of sight scenario (LOS), the lowest path loss recorded were between 105 and 115 dB, and the path loss recorded for highest transmitter–receiver displacement were around 115–130 dB. However, the averaged path loss swerved within 110–145 dB. The median path loss for this scenario is 123.0238 dB, and average path loss is 123.55 dB.
Table 1

Path loss measurement samples along with their average and fitted plots for body surface communication channel

In the non-line of sight scenario (NLOS), the path loss was contained within 80 and 105 dB for the lowest transmitter–receiver separation although for the highest separation it was amid 125–142 dB. The average path loss remained between 110 and 145 dB. The average median path loss in this band is approximately 135.58 dB. So, there is a variance of 10 dB between the LOS and NLOS scenarios. The average path loss for 3.4944 GHz for LOS and NLOS scenarios were established at 123.55 and 135.68 dB, respectively.

Likewise, in the 3.9936 GHz band for LOS setting, the minimum path loss swerved between 80 and 105 dB and maximum between 125 and 142 dB. The average path loss stays within 140 dB and surges monotonically from 110 dB. The median path loss now is 128.8984 dB. In NLOS scenario, the average path loss rises from 125 to 145 dB and the median is 144.9503 dB. The average path loss was 128.95 and 145.5152 dB, respectively, for LOS and NLOS scenarios.

The path loss for minimum transmitter–receiver separation in the 4.4928 GHz band stays between 110 and 118 dB, and for the maximum separation between transmitter and receiver, and the path loss stays between 125 and 140 dB. The average path loss is between 110 and 145 dB while the median path loss is 127.9846 dB for this band. However, in the NLOS scenario, the average path loss stays within 132 and 155 dB. The median path loss recorded in this band is 150.4428 dB. The average path loss was 129.03 and 151.0047 dB, respectively, for the LOS and NLOS scenarios.

In the 4.992 GHz LOS scenario, the minimum path loss is in the 100–105 dB range although the maximum is between 145 and 155 dB. The average path loss stays amid 100–145 dB, and the median path loss stays at 126.42 dB for LOS scenario. In NLOS scenario, the path loss minimum stays between 120 and 145 dB. The median path loss recorded is 141.2889 dB. The average path loss was 128.012 and 142 dB, respectively, for the LOS and NLOS scenarios.

The 5.4912 GHz band provides path loss in the range of 105–115 dB for the least separation and it increases up to 150 dB for the highest transmitter–receiver separation. The average stays in between 105 and 145 dB for the LOS scenario. The median path loss is 126.4260 dB. In NLOS conditions, the minimum stays at about 130 dB while the extreme path loss stays below 180 dB, and the average path loss logged is in the 130–150 dB range. The median path loss is almost 147.1812 dB. The average path loss recorded for the LOS and NLOS setups are 128 and 147.3485 dB, respectively.

The measured average and median path loss effects state that in body surface communication, the path loss difference between the LOS and NLOS scenarios can fluctuate from 10 to 15 dB for the various frequency bands and scenarios. Such hefty variations and path loss for small transmitter separations of about 1500 mm makes the body area channels extremely attenuating. These attenuations are a outcome of the absorption due to human body tissues.

6 Observations

The median path loss for the 30 diverse locations on the human body at the transmitter–receiver separation of 50–1500 mm with increments of 50 mm have been shown in Tables 1 and 2. The median path loss provides an insight into the attenuating environment of the human body and the shadowing in the body surface communication scenario.
Table 2

Median path loss for each frequency band for LOS scenario in the 3–6 GHz band for transmitter receiver separation of 50–1500 mm

Median path loss for the measured data for different transmitter receiver separation

Scenario-LOS

Distance (mm)

Frequency bands (in GHz)

3.4944

3.9936

4.4928

4.9921

5.4912

50

136.0143

116.8498

142.6309

131.7292

140.5642

100

143.6176

117.7677

146.9517

140.9434

144.3341

150

152.0886

133.0506

145.7136

145.8705

155.8577

200

151.1293

128.14

147.2132

150.785

155.3761

250

155.7058

146.288

149.8915

155.0827

148.0124

300

152.2303

145.7876

151.0709

154.981

145.8686

350

142.9543

147.4705

154.089

157.6821

148.1795

400

150.1371

159.2368

153.0748

156.0681

155.6873

450

149.0705

157.7497

153.4034

164.294

151.7098

500

145.7634

155.4324

157.8409

161.3984

163.3649

550

147.8018

154.3163

167.1929

156.4147

160.5457

600

151.3243

160.0143

155.5793

161.8024

160.7533

650

150.2506

154.527

151.0113

159.1876

162.3673

700

155.2289

145.9252

153.1594

157.0646

153.0811

750

152.7088

157.5373

159.5622

149.1324

149.1381

800

156.0042

158.56

159.3006

153.8256

149.5798

850

160.0153

157.6469

161.9585

157.456

154.4061

900

157.8649

160.3134

157.9905

158.3425

152.6504

950

152.7225

166.2534

159.2693

157.0002

165.4845

1000

156.7724

172.4768

162.4354

157.9831

163.0345

1050

153.9618

162.9533

160.0446

161.2745

164.7934

1100

159.489

171.5288

165.4528

160.5893

160.6547

1150

158.0489

172.4353

157.9787

161.7026

155.1509

1200

164.2238

161.9219

161.7486

166.1617

156.9944

1250

157.7709

159.6431

154.1837

159.2851

167.6649

1300

157.9299

171.9936

165.7723

161.2977

166.4682

1350

153.6251

167.3959

161.9773

161.0926

160.3333

1400

149.6335

168.5808

160.3802

164.0418

163.6075

1450

153.3251

163.3112

162.215

169.9287

168.4323

1500

153.925

164.9268

164.647

167.1659

165.9569

The median path loss for the 3–6 GHz band for the LOS scenario varies amid 80–145 dB for all the frequency bands. The overall path loss surges as the transmitter–receiver separation grows. The disparity in the median path loss can be attributed to shadowing that is triggered due to the human body. Similarly, in the NLOS scenario the median path loss fluctuates between 110 and 170 dB with an increase in transmitter–receiver separation. The effects of shadowing are more noticeable in the NLOS scenario for all the frequency bands. It is clearly observed from the Tables 1, 2 and 3 that path loss increases with distance, and thus the path loss model can be applied to it to calculate the path loss at any transmitter–receiver separation.
Table 3

Median path loss for each frequency band for NLOS scenario in the 3–6 GHz band for transmitter receiver separation of 50–1500 mm

Median path loss for the measured data for different transmitter receiver separation

Scenario-NLOS

Distance (mm)

Frequency bands (in GHz)

3.4944

3.9936

4.4928

4.9921

5.4912

50

148.5426

148.4194

140.3367

174.2956

165.942

100

151.2914

167.6778

163.8392

167.4007

177.8608

150

147.718

166.6841

174.072

167.8306

166.1807

200

151.78

166.7882

183.0136

163.8516

172.6681

250

152.7801

167.1571

181.2453

168.214

176.6839

300

155.3072

170.8456

181.3511

170.9839

176.5179

350

154.853

176.5553

181.836

171.5132

170.5656

400

168.5145

166.5087

185.0914

166.1097

173.0684

450

161.0028

170.111

178.3441

168.8241

166.4308

500

159.6543

165.4883

186.4462

172.3325

169.4396

550

165.2152

165.861

187.7767

171.0645

177.6784

600

162.6787

164.8171

186.8539

169.7387

170.9241

650

171.7887

179.1818

183.5335

166.3654

175.8162

700

162.5675

167.9104

177.9844

164.1408

183.7528

750

161.7335

170.5394

170.5283

170.2413

172.1376

800

164.6091

177.1849

171.6779

167.8074

172.3553

850

168.7024

178.4888

176.3061

165.2889

181.454

900

181.5487

178.2437

174.8892

174.9682

178.8744

950

169.0893

179.3365

173.7092

175.7674

178.374

1000

168.8731

174.8066

179.4808

183.5465

178.5932

1050

173.9595

172.6804

179.913

183.7019

178.8305

1100

169.7259

175.094

176.1345

176.5231

181.952

1150

167.5935

176.8818

182.1841

181.929

174.1827

1200

173.0511

176.946

182.7213

176.1864

174.5119

1250

165.9521

175.2819

186.5828

179.3094

179.7461

1300

163.6427

175.195

192.6379

182.5766

184.0251

1350

181.7679

185.1543

191.1218

183.8031

184.8732

1400

176.9071

187.6317

180.6546

177.418

187.2429

1450

172.1794

178.6853

177.1369

185.3016

184.7589

1500

168.5215

177.9766

180.231

170.4944

179.4595

The path loss after curve fitting the measured data can be expressed as the path loss Eq. (1) expressed as
$$PL\left( d \right) \, \left[ {\text{dB}} \right] \, = a \times log_{10} \left( d \right) \, + b + N$$
(1)
where a and b are the parameters of linear curve fitting N is a random variable that emulates shadowing and is conveyed by its standard deviation σ N . The values of a, b and σ N are given below in tabular form in Table 4 for reference.
Table 4

Parameters for equation 6.1, a, b and standard deviation σ N

Frequency bands (GHz)

Scenario

a

b

σ N

3.4944

LOS

10.14

− 363.7

21.4196

NLOS

20.64

− 278

18.8574

3.9936

LOS

37.2

− 357.3

19.4833

NLOS

16.65

20.58

8.7573

4.4928

LOS

14.93

− 31.3

10.8259

NLOS

15.31

− 342.6

21.2038

4.992

LOS

18.46

− 110.9

13.6931

NLOS

10.9

− 400.4

22.6649

5.4912

LOS

15.4

− 371.6

21.3682

NLOS

10.07

− 274.8

19.8444

Figure 3 shows a evaluation amid the fitted path loss for LOS and NLOS scenarios for all the frequency bands. The figure highlights difference between the LOS and NLOS scenarios, and the difference is more than 10 dB. The body surface to body surface communication scenario presents challenging channel conditions. In the case of the LOS scenario, a dominant LOS path greatly enhances the received power thereby decreasing the path loss. However, in the absence of such a path the path loss can reach levels of more than 150 dB for maximum transmitter–receiver separations.
Fig. 3

Fitted path loss with respect to distance for all frequency bands

7 Conclusion

The article presents channel models for ultra wideband-based wireless body area networks for the body surface to body surface communication scenario. The channel models are a outcome of more than 300,000 received power measurements in the 3–6 GHz band. The sub-band specific channel models are developed for transmitter receiver separation of up to 1500 mm. The models are for both line of sight and non-line of sight scenarios. The average difference in path loss amid the LOS and NLOS scenarios is more than 10 dB. In the absence of a LOS path between transmitter and the receiver, the path loss can increase to as much as 150 dB. These conclusions make the body surface scenario most challenging. The information collected about the channel characteristics of the UWB-based BAN channels will help scholars involved in the design of transmitters and receivers. The findings will be instrumental in determining the link budget.

References

  1. 1.
    Yuce, M. R. (2010). Implementation of wireless body area networks for healthcare systems. Sensors Actuators A Physical, 162(1), 116–129.MathSciNetCrossRefGoogle Scholar
  2. 2.
    Cao, H., Leung, V., Chow, C., & Chan, H. (2009). Enabling technologies for wireless body area networks: A survey and outlook. IEEE Communications Magazine, 47(12), 84–93.CrossRefGoogle Scholar
  3. 3.
    Astrin, A. W., Huan-Bang, L. I., & Kohno, R. (2009). Standardization for body area networks. IEICE Transactions on Communications, 92(2), 366–372.CrossRefGoogle Scholar
  4. 4.
    Astrin, A., et al. (2012). IEEE standard for local and metropolitan area networks part 15.6: Wireless body area networks: IEEE STD 802.15. 6-2012. Doc. is available IEEE Xplore.Google Scholar
  5. 5.
    Boulis, A., Smith, D., Miniutti, D., Libman, L., & Tselishchev, Y. (2012). Challenges in body area networks for healthcare: The MAC. IEEE Communications Magazine, 50(5), 100–106.CrossRefGoogle Scholar
  6. 6.
    Movassaghi, S., Abolhasan, M., Lipman, J., Smith, D., & Jamalipour, A. (2014). Wireless body area networks: A survey. IEEE Communications Surveys and Tutorials, 16(3), 1658–1686.CrossRefGoogle Scholar
  7. 7.
    Ghavami, M., Michael, L. B., & Kohno, R. (2004). Front matter. New York: Wiley.CrossRefGoogle Scholar
  8. 8.
    Siriwongpairat, W. P., & Liu, K. J. R. (2007). Ultra-wideband communications systems: Multiband OFDM approach. New York: Wiley.CrossRefGoogle Scholar
  9. 9.
    Porcino, D., & Hirt, W. (2003). Ultra-wideband radio technology: Potential and challenges ahead. IEEE Communications Magazine, 41(7), 66–74.CrossRefGoogle Scholar
  10. 10.
    Zasowski, T., Althaus, F., Stage, M., Wittneben, A., & Troste, G. (2003). UWB for noninvasive wireless body area networks: channel measurements and results. In 2003 IEEE conference on ultra wideband systems and technologies (pp. 285–289).Google Scholar
  11. 11.
    Allen, B., Ghavami, M., Armogida, A., & Aghvami, H. (2003). UWB technology. Communication Engineering, 1(5), 14–17.CrossRefGoogle Scholar
  12. 12.
    Mahesh, R. K. N., Ganesan, A., Kumar, M. P., & Paily, R. (2013). An ultra-wideband baseband transmitter design for wireless body area network. In VLSI design and test (pp. 26–34). Berlin: Springer.Google Scholar
  13. 13.
    Sawada, H., Aoyagi, T., Takada, J., Yazdandoost, K. Y., & Kohno, R. (2008). Channel models between body surface and wireless access point for UWB band. IEEE 802.15 WPAN Doc. IEEE 802.15-08-0576-00-0006 (pp. 1–14).Google Scholar
  14. 14.
    Molisch, A. F., Foerster, J. R., & Pendergrass, M. (2003). Channel models for ultrawideband personal area networks. IEEE Wireless Communications, 10(6), 14–21.CrossRefGoogle Scholar
  15. 15.
    Pang, Y., Lei, Q., Lin, J., Li, Z., & Ren, Y. (2013). Channel models of body area networks. Sensor Letters, 11(4), 731–735.CrossRefGoogle Scholar
  16. 16.
    Taparugssanagorn, A., Pomalaza-Ráez, C., Isola, A., Tesi, R., Hämäläinen, M., & Iinatti, J. (2009). UWB channel modeling for wireless body area networks in medical applications. In Proceedings of the international symposium on medical information and communication technology (ISMICT), 2009.Google Scholar
  17. 17.
    Yazdandoost, K. Y., Sayrafian-Pour, K., et al. (2009). Channel model for body area network (BAN). IEEE p802 15-08-0780-09-0006.Google Scholar
  18. 18.
    Cotton, S. L., D’Errico, R., & Oestges, C. (2014). A review of radio channel models for body centric communications. Radio Science, 49(6), 371–388.CrossRefGoogle Scholar
  19. 19.
    Smith, D. B., & Miniutti, D. (2012). Cooperative selection combining in body area networks: Switching rates in gamma fading. IEEE Wireless Communications Letters, 1(4), 284–287.CrossRefGoogle Scholar
  20. 20.
    Aoyagi, T. et al. (2008) Channel model for wearable and implantable WBANs. IEEE 802.15-08-0416-04-0006.Google Scholar
  21. 21.
    Seyed Mazloum, N. (2008). Body-coupled communications-experimental characterization, channel modeling and physical layer design. No. December, 2008.Google Scholar
  22. 22.
    Mohamed, M., Cheffena, M., Moldsvor, A., & Fontan, F. P. (2017). Physical-statistical channel model for off-body area network. IEEE Antennas and Wireless Propagation Letters, 16, 1516–1519.CrossRefGoogle Scholar
  23. 23.
    Cui, P.-F., Yu, Y., Lu, W.-J., Liu, Y., & Zhu, H.-B. (2017). Measurement and modeling of wireless off-body propagation characteristics under hospital environment at 6–8.5 GHz. IEEE Access.Google Scholar
  24. 24.
    Smith, D. B., & Miniutti, D. (2012). Cooperative body-area-communications: First and second-order statistics with decode-and-forward. In Wireless communications and networking conference (WCNC), 2012 IEEE, 2012 (pp. 689–693).Google Scholar
  25. 25.
    Smith, D. B., Miniutti, D., Lamahewa, T. A., & Hanlen, L. W. (2013). Propagation models for body-area networks: A survey and new outlook. IEEE Antennas and Propagation Magazine, 55(5), 97–117.CrossRefGoogle Scholar
  26. 26.
    Di Franco, F., Tachtatzis, C., Atkinson, R. C., Tinnirello, I., & Glover, I. A. (2015). Channel estimation and transmit power control in wireless body area networks. IET Wireless Sensor Systems, 5(1), 11–19.CrossRefGoogle Scholar
  27. 27.
    Alomainy, A., Hao, Y., Hu, X., Parini, C. G., & Hall, P. S. (2006). UWB on-body radio propagation and system modelling for wireless body-centric networks. IEE Proceedings-Communications, 153(1), 107–114.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Electronics and Telecommunication EngineeringCV Raman College of EngineeringBhubaneswarIndia
  2. 2.Department of Electrical EngineeringNational Institute of Technology RourkelaRourkelaIndia

Personalised recommendations