Power Efficient Communication for Medical Wireless Sensor Networks

Abstract

The thesis introduces a low complexity, power efficient communication system based on joint source-channel coding principle for medical wireless sensor networks. The design of the communication system relies on the separation theorem by C. Shannon where there exists a binary interface between the source coding and channel coding. Nevertheless, the well-structured attribute of the source coder is carefully taken into account when designing the channel coding part, and in return the channel quality is used to design the source coding. In particular, the source coder produces a binary source sequence which can be divided into different layers, and each layer requires a certain level of protection when transmitting over the channel. The unequal level of protection can be implemented either by allocating unequal power or by unequal transmission rate. To keep the source coder at low complexity level and to have scalable property, the compression efficiency of the source coder must be sacrificed when considering the source coding part alone. However, power gain for the whole system is obtained when the scalable property of the source coder is used to combat with wireless fading channels. Although the main work of the thesis is for the conventional Rayleigh fading channel, the advantages of the communication systems are still obtained for the human abdominal fading channel in medical wireless sensor networks. The low complexity, high robustness, and power efficiency of the communication systems proposed in this thesis are meaningful for medical wireless sensor networks.