A Cross-Layer Approach for Video Delivery over Wireless Video Sensor Networks

TL;DR

This paper introduces a cross-layer method for energy-efficient, reliable video delivery in wireless sensor networks, utilizing adaptive compression and a novel routing protocol to extend network lifetime and improve video quality.

Contribution

It presents a new adaptive video compression scheme combined with a novel routing protocol tailored for wireless video sensor networks, enhancing energy efficiency and video quality.

Findings

33% extension in network lifetime

3 dB improvement in video quality

Energy-efficient reliable routing protocol

Abstract

In this paper, we propose a novel cross-layer ap-proach for video delivery over Wireless Video Sensor Networks (WVSN)s. We adopt an energy efficient and adaptive video compression scheme dedicated to the WVSNs, based on the H.264/AVC video compression standard. The encoder operates using two modes. In the first mode, the nodes capture the scene following a low frame rate. When an event is detected, the encoder switches to the second mode with a higher frame rate and outputs two different types of macroblocks, referring to the region of interest and the background respectively. Furthermore, we propose an Energy and Queue Buffer Size Aware MMSPEED-based protocol for reliably and energy efficiently routing both regions towards the destination. Simulations results prove that the proposed approach is energy efficient and delivers good quality video streams. In addition, the proposed routing protocol EQBSA-MMSPEED outperforms its predecessors, the QBSA-MMSPEED and the MMSPEED, providing 33% of lifetime extension and 3 dBs of video quality enhancement.