A Hybrid PSO-GA Optimization Method for Relay Node Placement and Connection Point Detection in IoT Applications

With its enormous growth, the Internet of Things (IoT) expands its applicability by developing the living environment by introducing the smart home paradigm. Numerous wireless technologies are being investigated to support the rising interconnectedness of digital devices. Stability, lower cost, and reliable connections have made WiFi networks a dominant reliable connection mechanism in IoT in which one or more access points are installed to provide internet connectivity. Since the distance between IoT sensor nodes and the AP may exceed the communication range of nodes, relay nodes are placed in the network to reduce the connection distance. The problem of relay node placement in IoT to minimize the number of relay nodes is an NP-hard problem. Once the relay nodes have been placed, the connection point detection (CPD) problem plays a significant role in determining the quality of the network. CPD identifies which relay or access point each sensor node is connected to within its communication range. This study employs particle swarm optimization (PSO) for the relay node placement problem. In each PSO algorithm iteration, the genetic algorithm (GA) is used to solve the CPD problem to evaluate each particle. Our study is the first to take into account CPD in addition to the problem of relay node placement. It aims to minimize the number of relay nodes first and reduce the distance between sensor nodes to their associated relay nodes. Several experimental scenarios have been implemented to simulate the smart environment in IoT. The results indicate that the proposed method reduces the number of relay nodes by up to three and the average connection distance by up to 36% in comparison to other baseline approaches and a previous study.