doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 1 Project: IEEE Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Body area channel modeling for IEEE a] Date Submitted: [11Mar2004] Source: [Andrew Fort, Julien Ryckaert, Bert Gyselinckx] Company [IMEC] Address [Kapeldreef 75, Leuven, Belgium 3001] Voice:[+32(0) ], FAX: [+32(0) ], Re: [Channel model for communication around the body] Abstract:[Channel model for communication around the body] Purpose:[Contribute to channel modeling for body area applications] Notice:This document has been prepared to assist the IEEE It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release:The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 2 Outline Goal of our channel model Simulation results Proposed channel model Link Budget

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 3 Goal Channel model Determine required transmit power for a target BER as a function of the antenna position on the body, and the distance to walls or obstacles.

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 4 Propagation around the body through creeping wave EM waves propagate around the body via two paths: –Penetration (dielectric losses, tissues interfaces losses) –Creeping waves (diffraction mechanism) REMCOM XFDTD software together with a complete body model: Time step time step = 10ps

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 5 We determined the path loss near the human body by simulation. Exponential decay with angle difference Height difference less important Path loss is higher for higher frequencies Variance is larger in the interference region

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 6 Model with exponentially decaying loss Creeping wave propagation at 900MHz Breakpoint angle: Interferences between the clockwise wave, the counterclockwise wave and the penetrating wave Lower decay factor but larger variations

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 7 We propose a Rician Model to simulate nearby walls and obstacles Based on Rician line of sight channel model. The variance and attenuation of creeping wave << reflected paths. Ratio of Specular (Line of sight) power and Scattered (reflected power) must be estimated. Specular Component Scattered Components

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 8 The specular and scattered component powers can be estimated Specular component power can be estimated based on our simulated results. Scattered component power can be estimated based on the classical exponential path loss model: Parameters to be confirmed with simulations Path loss Path loss at reference distance (d o ) Reference distance (1 meter) Distance traveled by scattered components Path loss exponent 2

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 9 Rician Factor can now be estimated as a function of distance to obstacles Rician factor increases as we move further from obstacles Rician factor decreases with increasing carrier frequency Rician factor decreases with increasing angle separation

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 10 Link Budget : Best case occurs either very far or very close to obstacles. Scattered components begin to interfere with creeping wave. Scattered components dominates K << 0 dB Target BER = BPSK 900 MHz Required Transmit Power (dBm) Creeping wave dominates K >> 0 dB

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 11 Conclusions Creeping waves are a significant propagation mechanism affecting communication around the body. Reflected signal component from walls and obstacles in an indoor environment influence the required transmit power. We propose using a Rician model to perform a link budget as a function of antenna separation on the body and distance from obstacles.

doc.: IEEE /120r0 Submission March 2004 Bert Gyselinckx, IMECSlide 12 Future Work Improve estimates of P0 and P 0 for different antennas. Estimation of path loss exponent for different room geometries. Simulations to justify the Rician channel model and to compare with other distributions. UWB channel modeling.