Variants of VAMP for Signal Recovery in Wireless Sensor Networks

In wireless sensor networks (WSNs), compressed sensing (CS) is often used for gathering the signals of the sensors. Iterative recovery algorithms that retrieve the signals at the fusion center have to face the challenge of a non-uniform distribution of the receive powers induced by the individual sensors. This paper tackles the problem by a special form of bias compensation, which treats the biasing effects at the linear estimate individually. This consideration leads to variants of the well-known VAMP algorithm. We compare the performance of the new variants by numerical simulations in a WSN scenario.