Research On Data Processing Energy-Saving Mechanism In Wireless Sensor Networks

Wireless sensor network, as an emerging technology, has been widely applied in many fields, and has a great value which needs to be improved. The main purpose of wireless sensor network research is to maximize the network lifespan without compromising its service and designed function. Since power consumption is one of the major concerns affecting the network life cycle, how to build a power conservation mechanism into the network is the most significant area in the entire research. This thesis focuses on the data processing algorithms to reduce the amount of data to be transmitted through the network, and as a result to extend the sensor network service life. The background of this study is the application of wireless sensor network on smart garments. After the physical signals have been collected, we conduct the analysis on the human body physiology signals, such as ECG, breath, temperature value and other physiological parameters. Then data fusion is applied to combine measurement data from multiple channels to generate data packets which are transmitted in real-time from the patient to the remote medical center.Smart garment is designed to be worn by patient who requires real-time monitoring. It combines the technologies from electrical and biomedical engineering, and intelligent information processing, and makes the real-time communication between patient and medical care provider possible. However, this technology suffers from the same power supply problem as found in all mobile devices. Most of the energy is consumed in data transmission, so the network life largely depends on how much data needs to be sent. Physiology signals are weak signals and can be easily interfered during the collection by, such as power frequency disturbance, myo-electrical noise, and even human body movements which can lead to baseline offset. This study focuses on data fusion part of the entire network, with a specific emphasis on the extended Kalman filter to reduce the data and to save power. The real-time monitoring feature of the smart garment is achieved by live communication between the sensor network and the medical center. Any abnormal change of a patient's physiological signals can be received by patient care provider immediately.The algorithms and solutions proposed in this thesis have been carried on in terms of full theoretical analysis and the experimental verification. Data fusion enhances the energy conservation mechanism, and is validated for the sensor network. The solution proposed in this thesis provides valuable information for further exploration of the sensor network energy conservation research.