JoWUA

Abstract

The high rate of Internet of Things (IoT) applications is putting tremendous pressure on wireless sensor networks, which congest, raise latency, and packet loss because of insufficient computational and communication capability. This paper introduces a smart hybrid congestion control system based on fuzzy logic and particle swarm optimization (PSO) to improve network performance. Real-time inputs into the system, like packet loss and delay, are processed by the system through a fuzzy inference mechanism, and PSO dynamically decides the transmission intervals so that there is a better quality of service. The model is deployed on an ESP32 platform that comprises a hardware platform and real-time monitoring on the Blynk IoT environment and a Python-based analytical interface. The experimental analysis of the model in diverse network conditions proves that the hybrid model is much better than the traditional ones. The system obtains a small average delay of 6.33 ms as opposed to 24.96 ms for fuzzy-only and 22.09 ms for PSO-only models. The loss of packets is minimized to 0.30, which is better than fuzzy (0.6) and PSO (2.37) solutions. The congestion metric is kept at 0.97, which means that the network is stable. The hybrid approach also minimizes delay variation (jitter) to 1.9 ms, indicating better stability. Another result of the model is its quicker convergence (9.2 seconds) than PSO (18.5 seconds). The results substantiate that fast decision-making in fuzzy logic, coupled with adaptive optimization via PSO, leads to efficient congestion control. The design is a flexible, usable hybrid solution for real-time Internet of Things applications that require low-latency, reliable communication.