Hybrid Multi-Objective-Derived Horse Herd and Dragonfly Algorithm-Based Energy-Efficient Secured Routing in WSN

Abstract

Energy efficiency and security have become prominent aspects of Wireless Sensor Networks (WSNs) for transmitting data. The major challenge is to increase the “Quality of Service” (QoS) since it is restricted by fewer constrained resources. Although the existing routing protocols acquire efficient transmission, they are still subsisting with the challenges of improving security and energy conservation. It will be helpful in practical implications like the military, weather forecasting and healthcare industry. As the nodes are operated by the energy preservation of the battery, developing an energy-efficient protocol is challenging. Since WSN possesses massive numbers of sensor nodes, it also has to avoid long-term communication, and the clustering mechanism is a prerequisite. Considering the diverse objectives like delay, energy and distance, there are still remarkable challenges to develop the routing protocol. Moreover, security enhancement becomes another challenging issue over the network for data transmission through the sensor nodes. To overcome these issues, a novel energy-efficient routing model is designed in this paper by using a hybrid algorithm and the multi-objective derivatives. Initially, the clustering approach is employed to reduce the complexity and improve the security in an energy-efficient manner. Therefore, the Cluster Head (CH) is selected significantly using the novel Hybrid Horse Herd− Dragonfly Optimisation (HHHDO). The hybrid algorithm is developed by integrating Horse Herd Optimisation (HHO) and Dragonfly Algorithm (DA). Here, the optimal CH selection is achieved by deriving the multi-objective function regarding “distance, delay, energy, load and trust of nodes”. Trust is measured with the Neural Network (NN)-based HHHDO (NN− HHHDO) model. Depending on the trust evaluation and energy efficiency, the proposed model obtains better communication between the nodes. In the end, the experimentation of the recommended model is made using various measures and it’s ensured that the suggested technique enhances trust among the nodes and energy conservation in comparison to the existing algorithms.