Dynamic layered routing protocols based on BP-NN for underwater acoustic sensor networks

Abstract

Underwater acoustic sensor network (UASN) effectively improves the reliability of underwater long distance transmission through multi-hop transmission between multiple nodes. However, the limited energy of underwater acoustic nodes and large end-to-end transmission delay lead to uneven load and short lifetime of UASN. In this paper, we propose a dynamic layered routing protocol based on back propagation neural network (BP-NN), named as BPDLR, to address the above issues. In the proposed scheme, considering such factors as node topology, residual energy and communication environments, the UASN is vertically divided into the top layer, the middle layer and the bottom layer. The routing designs in each layer are different, and each of them is predicted by BP-NN training model for routing. In addition, we compare the proposed BPDLR protocol with the existing schemes such as static layered routing (SLR) protocol and depth-based routing (DBR) protocol in terms of network node survival, average end-to-end delay and packet loss rate. The simulation results show that, the proposed BPDLR protocol effectively alleviates the energy hole problem of UASN and extends the network lifetime.