Efficient routing based on past information to predict the future location for message passing in infrastructure-less opportunistic networks

Abstract

Opportunistic network (OppNet) is the recent evolution of mobile ad-hoc networks that has emerged as an active research subject in recent times. High mobility, frequent disconnections, sparse connectivity, no infrastructure, and limited resources are considered to be norms rather than hindrances. Hence, the challenges that one is likely to face while routing in opportunistic networks are very different from those faced in traditional wireless networks. Routing being the biggest challenge in such networks leads us to propose a new history-based prediction routing (HBPR) protocol for infrastructure-less OppNets which uses the behavioral information of the nodes to make predictions about their movements in the network. This helps to find and select a better next hop for the message to be routed to the destination. It also incorporates a method for the acknowledgment of the received messages, which helps in the buffer management of the intermediate nodes. Through simulations, the performance of HBPR is compared against popular solutions such as epidemic routing and Probabilistic routing protocol using history of encounters and transitivity, using the custom human mobility model. The HBPR performs fairly well in terms of number of messages delivered, average residual energy, overhead ratio, and average latency.