Improved Adaptive Fairness-Indexed Technique for Congestion Control in Underwater Networks

In underwater acoustic sensor networks (UASNs), effective congestion control is essential for ensuring reliable data transmission and resource efficiency, especially in the face of challenging environmental factors such as water velocity, pressure, temperature, and salinity. This paper introduces a novel Adaptive Fairness-Indexed Congestion Control (AFICC) mechanism that employs Q-learning to dynamically adjust transmission rates among nodes, ensuring equitable resource distribution and minimizing congestion. Our simulations, conducted with 100 nodes at depths of up to 1000 meters, reveal that AFICC significantly outperforms two existing algorithms: Rate-Based Congestion Control (RBCC) and Energy-Aware Congestion Control (EACC). Specifically, AFICC achieved a Packet Delivery Ratio (PDR) of 93%, compared to 85% for RBCC and 80% for EACC. Additionally, the End-to-End Delay was reduced to an average of 150 milliseconds with AFICC, while RBCC and EACC recorded delays of 220 milliseconds and 210 milliseconds, respectively. AFICC also demonstrated a 25% improvement in Energy Efficiency over EACC and maintained a Jain’s Fairness Index of 0.92, surpassing the 0.82 and 0.78 indices achieved by RBCC and EACC, respectively. These results confirm that AFICC offers a robust framework for congestion control in UASNs, effectively adapting to fluctuating underwater conditions while ensuring reliable communication and equitable resource allocation.