Investigating the ultra-wideband indoor wireless channel

Abstract

The indoor propagation channel appears differently to ultra-wideband (UWB) wireless systems than it does to narrowband (NB) sine wave systems. UWB impulses are short and generally do not overlap as multipath sine waves do. This paper presents the results of measurements made on the UWB indoor propagation channel. The channel impulse response (CIR) of the indoor office channel is extracted from the measurements using a variant of the CLEAN algorithm. The UWB propagation channel statistics are calculated from the CIRs. The analysis produces the following results. i) The pathloss versus distance obeys a power law (1/d/sup N/) curve fit, with the peak received power falling off with N=2.9 and the total received power falling off with N=2.1. This suggests that receiver architectures that can take advantage of the total received power can overcome much of the indoor channel pathloss. ii) A 4 element RAKE receiver shows a benefit as the peak power plus RAKE gain falls off with N=2.5, somewhat offsetting the excess pathloss of the channel. iii) The log-normally distributed standard deviation of the mean pathloss versus distance was found to be 4.75 dB, 4.04 dB, and 3.55 dB for the peak, peak plus RAKE, and total power pathloss, respectively. This suggests that a relatively small fading margin is required for UWB systems (compared to NB systems). iv) The RMS delay spread of the indoor channel increases with increasing T-R separation and also with increasing path loss. These results can aid in the design of UWB systems to overcome the challenges posed by the indoor channel.