A modified k-means-based cluster head selection and Philippine eagle optimization-based secure routing for MANET

Abstract

Decentralized mobile ad hoc networks (MANET) is a self-organizing and adaptive method comprised of multiple mobile nodes. Due to the dynamic and flexible nature of network topology, wireless communication links face major issues based on trust and security. The researchers came to know that MANET suffers greatly from the quality of service (QoS). So, MANET recommends the employment of an intrusion detection system to identify intrusions, namely attacks, message eavesdropping, and untrustworthy routing. Intrusion detection is considered a significant task to enhance network security from unauthorized access. The main issue in MANET is the limited battery source of sensor nodes which causes communication blockage after a certain transmission range. Therefore, to offer continuous data transmission without any intrusion, in this article, we introduced a secure high-yielding carry-out approach “modified k-means Philippine eagle (MKMPE) approach.” Initially, for the determination of the ideal cluster head (CH) from the cluster group, the modified k-means algorithm (MKM) employs for communicating information to the destination node without any packet loss. The CH is selected according to the values of credits namely indirect, recent trust, and direct trust. Then, with the support of Philippine eagle (PE) optimization, the energy-efficient and secure paths are discovered to transfer the information with minimum delay. To examine the potential ability of the proposed MKMPE approach, various state-of-the-art methods are compared with the suggested method in terms of evaluation metrics, namely energy consumption, network lifetime, end-to-end delay, latency, packet delivery ratio, throughput, and detection rate. This proposed MKMPE approach achieved a maximum detection rate of about 78%, a throughput rate of 0.86 bps, and minimum latency of 0.09 s concerning a maximum time interval of 45 ms.