Poster abstract: Ultra wideband biomedical wireless sensor neworks using wavelet lifting for image transmission

Image transmission in biomedical wireless sensor networks often requires low complexity, low power consumption with soft real-time constraint. However, the full separation of source coding and channel coding usually comes with large delay and/or complexity. To address these requirements, we propose a new direct mapping of source sample to channel sample that utilizes large bandwidth of ultra wideband (UWB). Our system combines a 5/3 integer lifting scheme wavelet transform followed by a pulse position modulation (PPM). The system's major advantages are low complexity, low power consumption, low processing delay and robustness to error at sensor node. Evaluation on an image transmission system over additive white Gaussian noise (AWGN) channel shows that our proposed system can achieve lossless or near lossless image with significant energy saving and very low delay.