Effects of Human Body on UWB Channel in Indoor Environments

Abstract

In this dissertation, the effects of human body on Ultra Wideband (UWB) channel in indoor environments are represented. Unlike previous communication system, UWB system has a large bandwidth. This leads to interference to the other communication systems in the same frequency bands. This feature makes UWB system deployable in line-of-sight (LOS) and slightly cluttered non-line-of-sight (NLOS) environment in which the signal undergoes less attenuation. In these environments, the UWB channel largely depends on surroundings of a transmitter (Tx) and receiver (Rx) antennas. In indoor environments, a human body is a major factor that changes channel characteristics. This dissertation dealt with the effect of human body on the UWB channel in indoor environments. First, this dissertation addresses UWB channel variation depending on the number of people in indoor LOS environments. To assess variation of UWB channels, four environments which have different room sizes and wall structures are considered. During measurements, people did not move around, but were just sitting on their chair with small motion if necessary. Because the UWB system operates in a wide bandwidth compared to previous communication systems, it is necessary to understand the frequency correlation characteristics of UWB channels. We found the correlation coefficients between two frequency tones with an interval of 10MHz are smaller than about 0.5. In the dissertation, we deal with a distance-dependent path-loss model, a frequency-dependent path-loss model, and time dispersion parameters. To provide a general channel model, we obtained the linear regression model with population density for each parameter. Next, the dissertation considered a situation where either LOS path is not blocked or slightly blocked by human bodies as a Rx is shifted by small-scale (1λ) distance while a Tx is fixed. In this situation, we measure the small-scale amplitude statistics in the absence and presence of human bodies and propose a statistical model of the small-scale fading distribution. From the measurement data, we found the best fitted channel model among several typical theoretical distribution models such as Lognormal, Nakagami, and Weibull distributions, showing good agreement with the empirical channel data. In the last part of dissertation, we dealt with the performance analysis of impulse radio (IR) UWB system based on the proposed small-scale fading distribution and also compare the performance with the existing channel model. Due to the fine time resolution of UWB system, the system mainly uses a rake receiver which consists of a number of correlators that are sampled at the delays related to specific number of multipath components. The dissertation considers two types of rake receiver, selective combining (SC) and partial combing (PC) rake receiver. The standard channel model, IEEE 802.15.4a, shows the best bit-error-rate (BER) performance. But this model does not include the effect of human body. When the effect of human body is included on 802.15.4a model, the BER performance is deteriorated.