On Providing Dynamic Route Reconfiguration in Congested MANET for Performance Amelioration Using Route Cache Trail (RCT) and Opinion Mining

摘要

The routing problems stemming in MANET is exacerbated by the unique characteristics of the network like dynamic topology, infrastructure less, absence of centralized authority, limited bandwidth, processing power, multi hop nature. The Multi Route AODV Ant routing algorithm (MRAA) implemented in this paper results from applying Ant colony optimization (ACO)/Swarm intelligence techniques to AODV routing protocol. This proposed protocol namely MRAA is intelligent enough to balance the load in the Densely Routed Path (DRP)/Core Forwarding List (CFL)/ Most preferred route (MPR). The ant's normal behavior of constructing an optimal and promising path in search of food from the nest is mimicked in this network. This behavior emulation in network is sure to short circuit/bypass routing bottlenecks that causes network diversion, impairment and outage. The selection of obstacle/hurdle free path is sure to attract heavy network traffic to it amplifying congestion. Load balancing in DRP necessitates/demands dynamic route reconfiguration/reconstruction to eventually reach the destination. This route reconfiguration/recomputation adds significant performance burden in timely sensing of destination node. The node mobility coupled with this compounds the problem of realizing an optimal routing path. Performance Amelioration through Pheromone based Dynamic Route Reconfiguration (DRR) technique is used to detect/find secondary forwarding/routing list to successfully reach the destination. ROUTE CACHE TRAIL (RCT) a performance metric in synchronization with other measures like opinion mining is used to map the nodes in DRP to the alternate Sparse Populated Route leading to the same destination. RCT an intuitive process of storing/holding the latest/currently used routing path with its associated parameters viz. source id, RREQ, sequence no., Broadcast Id, Destination Id etc. in all the participating nodes for future routing endeavors and helps in backtracking the RREP packets to the source. Opinion Mining of the comprehensive routing information generated from the network enables the core node in the primary path in timely evaluation of the incoming node as genuine or not attesting to its participation in the alternate route creation. The volatile nature of RCT mirroring the varying pheromone concentration on the path provides scope for imbibing dynamicity in the routing process.