Optimal relay node selection in time-varying IoT networks using apriori contact pattern information

Abstract

Mobility of computing devices in Internet of Things brings the challenge of robust data forwarding over time-varying networks. To realize robust data forwarding methods in time-varying IoT networks, relay nodes need to be selected at every instant in time to improve the QoS in such networks. In this work, we propose a method for online relay node selection by utilizing the partial knowledge of apriori network contact patterns. The network contact pattern information is generally obtained by various machine learning and prediction methods. The proposed method selects a relay node based on joint optimization of two network parameters namely, data latency and link reliability. A heuristic cost function is modelled to jointly optimize the data latency and link reliability utilizing the apriori contact pattern information. By minimizing the cost function, an optimal relay node is chosen at every instant in time. During performance evaluation, network contact patterns of IoT devices are modelled using the homogeneous Poisson point processes. The contact period information of all the IoT devices with their neighbours is updated continuously and a relay node is found in an online manner. Simulation results indicate that the proposed method significantly improves data latency and the reliability of links when the knowledge of apriori contact patterns of IoT devices is utilized. Performance of the proposed data forwarding method is analysed in terms of transmission range, mobility tolerance, and connectivity parameters of time-varying IoT networks. The proposed method indicates additional gain in terms of packet replication cost when compared to the conventional methods.