Efficient digital encoding and estimation of noisy signals

Abstract : In many applications in science and engineering one must rely on coarsely quantized and often unreliable noisy measurements in order to accurately and reliably estimate quantities of interest. This scenario arises, for instance, in distributed wireless sensor networks where measurements made at remote sensors need to be fused at a host site in order to decipher an information-bearing signal. Resources such as bandwidth, power, and hardware are usually limited and shared across the network. Consequently, each sensor may be severely constrained in the amount of information it can communicate to the host and the complexity of the processing it can perform. In this thesis, we develop a versatile framework for designing low-complexity algorithms for efficient digital encoding of the measurements at each sensor, and for accurate signal estimation from these encodings at the host. We show that the use of a properly designed and often easily implemented control input added prior to signal quantization can significantly enhance overall system performance. In particular, efficient estimators can be constructed and used with optimized pseudo-noise, deterministic, and feedback-based control inputs, resulting in a hierarchy of practical systems with very attractive performance-complexity characteristics.