On-demand Routing Protocol Based on Prediction Model in Dynamic Scene
Mengqian Xu, Jin Li, Xiaoyu Dong, Ruimeng Gan, Zhe Xiao · 2023
Mobility is a pervasive phenomenon in wireless sensor networks (WSNs), which can significantly impact network routing performance. By combining trajectory prediction with appropriate routing protocols, the problem of high network latency caused by node mobility can be effectively solved. This paper proposes a real-time routing protocol for WSNs that leverages an intelligent trajectory prediction method based on Long Short-Term Memory (LSTM) algorithm. Our model uses the node's historical trajectory and the memory of neighboring nodes to predict the network topology to recalculate the route in dynamic scenes. Simulation and experimental results demonstrate that our proposed protocol outperforms existing protocols in terms of packet delivery ratio and end-to-end delay.