Optimal Wireless Communication Design over Impulsive Noise Environment

Abstract

The wireless network has been sharply developed to support wireless service such as message, video and so on. Such network systems are designed to achieve optimal performance in an additive white Gaussian noise (AWGN) environment. There are, however, some applications where additive non-Gaussian noise or impulsive noise is dominant. Since the existing receivers no longer provide the optimal performance in such an environment, it is important to design communication receivers based on a proper impulsive noise model to be optimal. Among various schemes to reduce impact of impulsive noise, methods using thresholds have been used so that the overall system guarantees low complexity. First, nonlinearity preprocessors with threshold has been popularly used. The primary function of the preprocessor is to reduce impact of the received samples interfered with impulsive noise, whereas the samples corrupted with Gaussian noise is not preprocessed. There are two preprocessors: blanker and soft-limiter that blocks out or clips out the received signals with a larger amplitude than the predetermined threshold. Blanker and soft-limiter are required to determine the optimal blanking and clipping thresholds for best performance. However, the blanking or clipping threshold is not calculated on an on-demand basis in practical systems. Second, adaptive modulation with threshold to satisfy target error rate is one of the efficient method for communication system. An adaptive modulation is a promising scheme to improve spectral efficiency by switching modulation order by a pre-determined SNR threshold. The basic idea of adaptive modulation is an instantaneous adaptation for link-budget in flat fading through controlling transmission power, symbol transmission rate, error rate, coding rate and scheme, and so on. Under this adaptation, the scheme offers a higher link spectral efficiency without wasting power or degrading performance of the system. However, there is little research outcome of adaptive modulation in impulsive noise environment. The main objective of this dissertation is to propose simple calculation of thresholds for the nonlinearity preprocessors for practical communication system. These simple and neat expressions allow the receiver to calculate or update the blanking or clipping threshold of the preprocessor on the fly when it is necessary, which may be an essential functionality for practical receivers. Also, since performing channel estimation and updating the nonlinearity preprocessor based on instantaneous fading channel variation in real-time may be difficult in practical systems, this dissertation considers a soft-limiter with the threshold that is computed only by average channel power, not instantaneous channel state, as the preprocessor to maximize a power ratio between a received signal and only noise. Then, bit error rate (BER) performance of the receivers with various practical nonlinearity preprocessors in an impulsive noise environment is given as closed forms. In addition, thresholds for adaptive modulation in impulsive noise environment is proposed. To evaluate the communication system with the adaptive modulation, performance analysis of average spectral efficiency is performed.