Robust Image Transmission Over Wireless Sensor Networks

Robust image and video communications have become more imperative due to the ubiquitous proliferation of multimedia applications over wireless sensor networks. In this work, the transmission distortions on the image data induced by both channel and instant node failures for Wireless Sensor Networks (WSN) are considered. The effect of two techniques and their integration with multipath transmission are investigated to compensate the multimedia distortions at the expense of incurring additional energy consumption and/or wasting bandwidth resources. First technique is watermarking based error concealment utilizing discrete wavelet transform for embedding downsized replicas of original image into itself. The other is conventional Reed-Solomon (RS) coding utilizing additional information bits to correct bit/symbol errors. Performance results obtained from extensive simulations utilizing a communication and energy model applicable to WSN show that error concealment (EC) integrated schemes, especially EC with multipath fusion (ECMF), are more promising to compensate losses than RS-coding-integrated and pure multipath transmission techniques in WSN.