Bootstrap Aggregated Mutual Dependency Ensemble Clustering and Learning Agent Based Approach to Eliminate Stale Routes in MANET

Abstract

Mobile Ad hoc networks deploy the network with the support of self-organizing and self-configuring mobile nodes. Due to the lack of centralization, the topological structure of the network fluctuates frequently. Preserving stable link communication to obtain reliable data transmission is the key challenge in the dynamic wireless network environment. This stimulates discrepancy on discovered route paths. To address this issue a novel approach called bootstrap aggregated mutual dependency ensemble clustering and learning agent based approach to eliminate stale routes in MANET (BAMDEC-LABA) is introduced. This algorithm is used to identify the stable link based on the metrics such as residual energy, receiving signal strength, less hop count and node behavior. Maximum dependency with less hop count route paths are classified by employing bootstrap aggregation method. Learning agent examines the node behavior and identifies the selfish and corruptive nodes using node cooperativeness and trust value. The occurrence of the link failure due to the malicious nodes intimated to all the nodes with the distribution of route error packet. The inconsistent route path is eliminated from the cache to preserve the link failure. The performance of the proposed approach is evaluated with different performance metrics such as routing overhead, packet delivery ratio, packet drop rate, and delay. When compared to state-of-the-art approaches, the proposed BAMDEC-LABA technique on an average minimizes the routing overhead by 26%, improves the packet delivery ratio by 18%, packet drop rate is considerably reduced by 68% and delay is found to be minimized by 27%. The proposed method outperforms when compared to state-of-the-art approaches.