Distributed Recovery of a Gaussian Source in Interference with Successive Lattice Processing

A scheme for recovery of a signal by distributed listeners in the presence of Gaussian interference is constructed by exhausting an “iterative power reduction” property. An upper bound for the scheme’s achieved mean-squared-error distortion is derived. The strategy exposes a parameter search problem, which, when solved, causes the scheme to outperform others of its kind. Performance of a blocklength-one scheme is simulated and is seen to improve over plain source coding without compression in the presence of many interferers, and experiences less outages over ensembles of channels.