Location based Routing in Opportunistic Networks using Cascade Learning

Abstract

As a key enabling technology for pervasive computing, Opportunity Network has recently attracted a great deal of research work in machine learning domain. In this paper, we present a new opportunistic network routing strategy that tries to send messages to a destination location area rather than a single node. To improve routing in Opportunistic networks, we apply cascade learning (a type of integration-based machine learning). We demonstrate that the proposed Cascade Learning based Routing Protocol (CLRP), outperforms existing machine learning-based protocols LOOP and MLProph on various performance metrics, including delivery probability and average latency, using real data mobility traces in simulation.