Trust-based neighbor selection using activation function for secure routing in wireless sensor networks

Abstract

Wireless sensor network (WSN) nodes rely on their neighbors to transmit the sensed information to a sink node. A transmitting node assumes that its neighbor is secure and relays its information in an optimal manner. A network’s wireless communication medium and decentralized nature retards its performance by exposing it to internal and external attacks. Therefore, secure networks are placed and vital in selecting trusted neighbors to secure transmission. Trust models and the reputation management system takes advantage of the network’s limitation, resulting in the weekly node detection of malicious neighbors and inconsistent network performance. This paper proposes an activation function-based trusted neighbor selection (AF-TNS) for resource-constrained WSNs to enhance network security. AF-TNS works in two phases: trust evaluation with energy constraint and additive metric-based node evaluation to retain the trustworthiness of the neighbors. The random transigmoid function employed simplifies the complex decision-making process of the AF by distinguishing trusted and un-trusted node to retain network performance. Simulation results show that AF-TNS improves network performance by improving the malicious detection rate and retaining the network’s lifetime. From the experimental result, AF-TNS method ensures minimum delay (8.5 s), minimum energy (8.53 J), high throughput (149 kbs), high network lifetime (390 s), also have less false detective rate (1.5%) while transmitting the network information.