Reliable multimedia transmission over wireless sensor network

Abstract

Nowadays, video streaming application is widely used in wired as well as wireless environment. Extending this application into Wireless Sensor Networks (WSN) for IEEE 802.15.4 network has attracted lots of attention in the research community. Reliable data transmission is one of the most important requirements in WSN especially for multimedia application. Moreover, multimedia application requires high bandwidth and consumes large memory size in order to send video data that requires small end-to-end (ETE) delay. To overcome this problem, rate control serves as an important technique to control the bit rate of encoded video for transmission over a channel of limited bandwidth and low data rate which is 250kbps with small Maximum Transmission Unit (MTU) size of 127 bytes. Therefore, a rate control model called enhanced Video Motion Classification based (e-ViMoC) model using an optimal combination of parameter setting is proposed in this thesis. Another challenging task to maintain the video quality is the design of an enhanced transport protocol. Standard transport protocols cannot be directly applied in WSN specifically, but some modifications are required. Therefore, to achieve high reliability of video transmission, the advantages of User Datagram Protocol (UDP) features are applied to the proposed transport protocol called Lightweight Reliable Transport Protocol (LRTP) to tailor to the low data rate requirement of WSN. Besides, priority queue scheme is adopted to reduce the end-to-end delay while maintaining the reliability and energy efficiency. Evalvid simulation tool and exhaustive search method are used to determine optimal combination of quantization scale (q), frame rate (r) and Group of Picture (GOP) size (l) values to control the bit rate at the video encoder. The model of e-ViMoC is verified both with simulation and experimental work. The proposed transport protocol has been successfully studied and verified through simulation using Network Simulator 2 (NS-2). From the simulation results, the proposed e-ViMoC encoded video enhances the Packet Delivery Ratio (PDR) by 5.14%, reduces the energy consumed by 16.37%, improves the Peak Signal to Noise Ratio (PSNR) by 4.37% and reduces the ETE delay by 23.69% in average, compared to non-optimized encoded video. The tested experiment experiences slightly different result where the PDR is 6% lower than simulation results. Meanwhile, the combination of e-ViMoC and LRTP outperforms the standard transport protocol by average improvement of 142.9% for PDR, average reduction of 8.87% for energy consumption, average improvement of 4.1% for PSNR, and average reduction of 19.38% for ETE delay. Thus, the simulation results show that the combination of proposed e-ViMoC and LRTP provides better reliability performance in terms of the PDR while simultaneously improves the energy efficiency, the video quality and ETE delay.