DLCOF: A DEEP LEARNING–INFUSED CROSS-LAYER OPTIMIZATION FRAMEWORK WITH FUZZY LOGIC FOR ENERGY-EFFICIENT ROUTING IN WIRELESS NETWORKS

Abstract

The wireless networks require dynamic routing systems that can maintain energy efficiency, reliability, and spectrum usage. This is typically not the case in traditional routing strategies because the strategies are usually configured at the static level and have inadequate coordination across the layers. The proposed structure combines deep learning-enhanced cross-layer optimization with fuzzy logic in order to develop smart routing choices in physical, data link, and network layers. Deep learning algorithms forecast the traffic changes, traffic mobility, and network conditions, allowing to make advance route decisions, and the fuzzy logic addresses the uncertainties in signal density, energy, and congestion indicators. Balanced power usage, lower end-to-end latency, and better delivery of packets are guaranteed by the hybrid optimization engine. The simulation findings show that it has high improvements compared to other protocols with 98% ratio of packet delivery, 90.4 ms latency, 1.5% of collision and 500.1 J energy usage per 10 nodes. This integration will provide a strong spectrum management and it will be very stable even when in dense network conditions. It develops a flexible and dynamic routing paradigm that maximizes the spectral efficiency, reduce data collisions, and increase lifespan of the network in the next generation wireless setting.