Intelligent Routing Using Time Series: CF – Prophet

Abstract

Opportunistic network is an extension of Adhoc network, subclass of Delay Tolerant Network and derivation of Mobile Adhoc Networks. The main characteristic is an intermittent connectivity resulting in indefinite waiting time for path establishment due to which routing is a tedious task. Prophet improves the network performance but it does not resolve congestion and flow control issues which arises due to adverse choice of traditional nodes in routing mechanism. Many researchers proposed variants of Prophet but encountering the congestion and flow issues into account have not been proposed yet. We proposed a novel approach called CF – Prophet using time-series analysis wherein packet count taken sequentially and fits those into historical data for predicting the scope of congestion and flow control issues. We simulate and analyze the proposed approach by availing the network performance metrics i.e. Communication transmission ratio, Dormancy Ratio and Steering Dormancy Ratio.