Research on AdHoc network routing algorithm for link stability prediction

Abstract

Abstract This study aims to improve network communication efficiency and reliability by researching routing algorithms for predicting link stability in AdHo networks. AdHoc network is a self-organizing, decentralized network structure in which nodes establish temporary connections through wireless communication, making the network resilient and flexible. However, due to the continuous changes in network topology and the instability of wireless links, routing problems in AdHoc networks have become particularly complex. This article proposes a backup path routing algorithm LS-BPR based on link stability prediction. This mechanism takes the path with the minimum delay as the main path, uses the cost function as the standard for selecting backup paths, and then uses the ARIMA model to predict links with good link stability as backup paths. We simulate and compare LS-BPRAODV and AODV protocols using the QualNet simulation platform. The simulation results show that the LS-BPRAODV, with the addition of a prediction mechanism, improves network performance, reduces the number of source node routing discoveries, reduces the number of route breaks, significantly reduces the packet loss rate and information reception delay in the network, and improves the average throughput of the network. This study provides new theoretical and empirical support for the routing algorithm for link stability prediction in AdHo networks, which is of great significance for promoting the development of wireless self-organizing networks.