Performance optimization of multimedia services in wireless multimedia sensor networks (WMSN) at the socket level

Abstract

Wireless Multimedia Sensor Networks (WMSNs) can collect video, audio, and scalar data, making them highly valuable for Internet of Things (IoT) applications. There is a growing demand for improved Quality of Service (QoS) in WMSNs, particularly in terms of energy efficiency, latency, and reliability. Meeting these requirements poses significant challenges, as researchers must develop methodologies that optimize resource utilization, enhance QoS and network performance, and minimize packet loss. QoS enhancements can be achieved at various layers of a WMSN from the Application layer down to the Physical layer. In this research, we implement priority schemes tailored for high data rate scenarios; packet traffic follows burst arrival and departure patterns. Our results show that assigning dedicated sockets on a single channel and prioritizing video and voice packets can significantly improve the QoS of the network. The simulation results clearly indicate that the single-channel sockets model outperforms the multichannel model across key performance metrics. In terms of throughput, the single-channel sockets scheme achieves approximately 10–12% higher values, showing its efficiency in handling increased traffic flows. Delay performance is also significantly improved, with about 15–20% lower latency under heavy load compared to the multichannel design. Furthermore, packet loss is reduced by nearly 10% in the initial phase, stabilizing at levels close to zero, thereby highlighting the robustness of the proposed single-channel sockets approach in supporting resilient and high-quality multimedia communication.