A thesis on Coding Techniques for Analog Wireless Communication System with Feedback

Abstract

Analog feedback communication system (AFCS) is a promising communication system especially for power constrained devices. AFCS uses simple transmitter architecture comprising of an adaptive pulse amplitude modulator (PAM) in the forward channel. The modulation index of this adaptive PAM is adjusted on the basis of the feedback of the estimate received from the receiver. Receiver uses Bayesian technique to estimate the transmitted symbols. This iterative feedback algorithm reduces the mean square error (MSE) of estimation during each iteration. MSE is the single criterion used to optimize both transmitter and receiver. This AFCS architecture is known to offer capacity reaching to the Shannon boundary. The absence of coding decoding requirement makes the system simple, cost effective and power efficient. Moreover, in the case of changing channel conditions, AFCS has an edge over digital systems due to its adaptive nature. As such, complete redesign of the system due to change in the channel condition is not needed. Despite many advantages and re-initiation of research in AFCS, the system is not yet realized and not in use. Still, various aspects of AFCS are being investigated.