Outlier detection based energy efficient and reliable routing protocol using deep learning algorithm

Abstract Wireless sensor network have also played a vital role in the observation and management of agricultural land terms climate, water usage, crops, etc. Due to open communication system low battery power sensors, sector still faces issues with energy consumption, information forwarding, privacy. Thus, an energy‐efficient routing during transmission WSN‐based smart agriculture is suggested this study applying feed‐forward neural detect outliers. Outlier identification, CH‐selection, Relay Node (RN) selection are three phases method. detection performed deployed nodes for categorises attack from normal nodes. CH‐selection using chaotic moth‐flame optimization technique according distance, node degree, centrality factor residual level, these parameters determine which will become Cluster Head. Then reliable protocol designed NB‐based probability method RN selection. MATLAB software used test proposed Detection based Energy Efficient Reliable Routing Protocol verify its performance. The effectiveness proposed‐model tested some prior wireless protocols environment‐fusion multipath protocol, dynamic Multi‐hop Protocol, SEMantic CLustering, efficient protocol. algorithm attained 0.91 (%)Packet Delivery ratio, 0.08% packet loss, 0.91% Average energy, 2.8 (Mbps) throughput, 26 (sec) Delay.