An Optimal Strategy for Cluster Hop Detection for Delay Tolerant Networks

Abstract

Improvements over the strategies for making the delay tolerant networks (DTN) more effective are always been the focus for the researchers due to the continuously increasing demands of DTNs from various domains of applications ranging from tactical to entertainment to education and business. In the last few years, the increase in usage has significantly increased the challenges for building more smart applications based on DTN. From the parallel research outcomes, it can be observed that the further improvements are bottlenecks due to primarily two facts. Firstly, due to the fundamental nature of the DTNs, the higher distance between the intermediate nodes causes huge delay in forming the clusters to transmit data with multiple hops. Secondly, the detection of the cluster members is purely based on the location or proximity to the source or base nodes and due to this, during the actual transmission of the data, nearly 20% of the time is dedicated to replicate the data in the intermediate nodes. Thus, as a result, the delay in the network increases further. Henceforth, this work proposes a novel cluster member detection strategy based on data affinity with multi-level queue for interim data storage. Due to this strategy, two major improvements can be observed. Firstly, the identification of the cluster members is now much more effective and secondly, the delay in the transmission can be further reduced with a great extent. During the simulation, the proposed work demonstrated the average cluster node detection time as 0.0146 sec with a significant 90% reduction of memory utilization and with a cache allocation accuracy of 85%, compared with the parallel research outcomes.