Improved Optimization and Long Short Term Memory‐Based Energy Prediction for Cluster Head‐Based Routing in Wireless Sensor Network: Reliable Transmission and Data Perseverance

ABSTRACT The efficient cluster‐based routing strategies are essential because of inherent limitations of wireless sensor networks (WSNs) such as limited bandwidth and energy resources. In WSN, the utilization of energy can be reduced by organizing nodes into clusters; thus, it prolongs the lifespan of the network. Additionally, it is essential to maintain data reliability and accuracy in environments prone to interference by utilizing reliable transmission strategies. Thus, a novel approach, namely, self‐improved secretary bird optimization algorithm for cluster head optimization and reliable transmission (SISBOA‐CORT), is proposed to address such issues. This approach encompasses two essential phases: optimal cluster head (CH) selection‐based routing and reliable transmission to ensure efficient cluster‐based routing and perform data transmission in a reliable manner. In the optimal CH selection‐based routing phase, the nodes in WSN are grouped into clusters. In each cluster, an optimal CH is selected using the self‐improved secretary bird optimization (SISBO) algorithm under consideration of constraints such as trust, link quality, latency, energy, and distance. An LSTM‐based energy prediction model is employed in this approach to predict the energy levels of each node in the network. For the reliable transmission phase, Manhattan distance is employed to find the shortest distance between the source node and the destination node. The efficiency of the SISBOA‐CORT approach is validated experimentally in terms of metrics such as delay, trust, RSSI, etc. The SISBOA‐CORT achieved a delay rate of 1.187 s, a residual energy value of 3.097 J, and an RSSI of −139.265.