Abstract

This paper presents the effect of a human body on ultra-wideband off-body wireless body area network radio channels. The work is based on static measurements in an anechoic chamber by using a vector network analyzer in a 2—8 GHz frequency band. Thirteen antenna locations on a human body are used while the off-body node is attached to a pole at a distance 1 and 2 m from the test subject. Two planar prototype antennas are applied: dipole and double loop. The data analysis is carried out in time domain by observing the first arriving signal components of the channel impulse responses. The classical path loss model fitting results in path loss exponents of 1.7 and 1.4 for the dipole and double loop, respectively. The classical path loss model is not found to be suitable in all cases in wireless body area networks as the path loss exponent varies greatly depending on the antenna site under examination. The absolute path losses reach values between 50.6…66.5 dB (dipole) and 49.9…68.2 dB (double loop) depending on the antenna location. In most of the cases, the double loop performs better than the dipole. When averaged over all antenna sites, the mean path losses lie in the range of 57.6…61.3 dB, and their standard deviation is approximately 4…5 dB depending on the distance and antenna focus.