This thesis studies digital compression of sparse signals via information-theoretic limits and compressed sensing algorithms. A rate-distortion function with multiple constraints is proposed and studied for sparse sources. Two different compressed sensing problems with scalar quantization are considered. First, Bayesian approximate message passing algorithms are applied to single and multi-terminal settings. Second, uniform approximation guarantees are derived for distributed one-bit compressed sensing.
«
This thesis studies digital compression of sparse signals via information-theoretic limits and compressed sensing algorithms. A rate-distortion function with multiple constraints is proposed and studied for sparse sources. Two different compressed sensing problems with scalar quantization are considered. First, Bayesian approximate message passing algorithms are applied to single and multi-terminal settings. Second, uniform approximation guarantees are derived for distributed one-bit compressed...
»