Secure Data Transmission in Wireless Sensor Networks via Hybrid LSTM‐FCM Deep Learning and Enhanced LEACH‐C Protocol (FLLC)

ABSTRACT Wireless sensor networks (WSNs) are vulnerable to serious security risks because of their open communication settings, decentralized architecture, and resource constraints. To address these challenges, a novel Deep Learning Fuzzy‐based Trust‐aware Routing Protocol (DLF‐TRP) is proposed. This model integrates long short‐term memory (LSTM) networks and fuzzy C‐means clustering to enhance node‐ and path‐level security. The LSTM module monitors node behavior and assigns anomaly scores, which are processed using fuzzy C‐means clustering to classify nodes into trust‐based groups. This fuzzy‐based trust evaluation allows for dynamic and context‐aware trust score assignment, accommodating uncertainty and partial information in WSN environments. DLF‐TRP performs intelligent path selection by favoring routes with high‐trust nodes while avoiding suspicious or potentially malicious ones. The proposed FLLC (fuzzy C‐means + LSTM + enhanced LEACH‐C) routing model enhances secure and energy‐efficient data transmission in WSNs. Simulation results show a 37.8% increase in network lifetime, 7.65% improvement in PDR, and 14.8% reduction in delay compared to existing DL‐GMA and EEDLCR protocols. The model ensures adaptive trust evaluation and efficient cluster‐head selection for robust performance.