An Unequal Clustering and Multi-Hop Routing Protocol Based on Fuzzy Logic and Q-Learning in WSNs

Clustering-based routing techniques are key to significantly extending the lifetime of wireless sensor networks (WSNs). However, these approaches often do not address the common hotspot issue in multi-hop WSNs. To overcome this challenge and enhance network lifespan, this study presents FQ-UCR, a hybrid approach that merges unequal clustering based on fuzzy logic (FL) with routing optimized through Q-learning. In FQ-UCR, a tentative CH employs a fuzzy inference system (FIS) to compute its probability of being selected as the final CH. By using the Q-learning algorithm, the best forwarding cluster head (CH) is chosen to construct the data transmission route between the CHs and the base station (BS). The approach is extensively evaluated and compared with protocols like EEUC and CHEF. Simulation results demonstrate that FQ-UCR improves energy efficiency across all nodes, significantly extends network lifetime, and effectively alleviates the hotspot issue.