A joint source-channel-network coding design for Wireless Sensor Network

Abstract

Wireless Sensor Network (WSN) data have correlations in between, since sensor nodes are usually monitoring a specific physical environment, which shares common natural features. The redundancies in sensor network data can be removed by the cooperative compressed sensing (CS). In this paper, a cross-layer communication framework which is able to accommodate joint source-channel-network coding through cooperative CS is proposed. We consider a hierarchical WSN, in which wireless relay nodes serve exclusively as forwarders. Analog transmission is assumed that analog network coding (ANC) can be performed at the relays. ANC is demonstrated to have more transmission capacity than its counterpart digital network coding. Joint source-channel coding is performed at the each sensor node by traditional single CS to compress data in temporal dimension, while delivering data in an ANC way is to compress aggregated network data in spatial dimension by cooperative CS. This joint source-channel-network coding scheme is able to explore both intra-signal and inter-signal correlation to sample, compress and transmit WSN data by an attractive power efficient way.