Compressed Sensing based Protocol for Efficient Reconstruction of Sparse Superimposed Data in a Multi-Hop Wireless Sensor Network

Abstract: We consider a multi-hop wireless sensor network that measures sparse events and propose a simple forwarding protocol based on Compressed Sensing (CS) which does not need any sophisticated Media Access Control (MAC) scheduling, neither a routing protocol, thereby making significant overhead and energy savings. By means of flooding, multiple packets with different superimposed measurements are received simultaneously at any node. Thanks to our protocol, each node is able to recover each measurement and forward it while avoiding cycles. Numerical results show that our protocol achieves close to zero reconstruction errors at the sink, while greatly reducing overhead. This initial research reveals a new and promising approach to protocol design through CS for wireless mesh and sensor networks.