Research and Application of Median Filtering Algorithm for Wireless Sensore Network
|School||Central South University|
|Course||Control Science and Engineering|
|Keywords||Wireless sensor networks Signal processing Median filter Noise suppression Frequency Selection|
Wireless sensor network information processing, digital filtering is very important in information processing a data processing technology. Limited power wireless sensor nodes, compute, storage capacity, digital filtering method for wireless sensor networks must meet basic performance requirements on the basis of, with simple, efficient, and other features. Median filter is a wireless sensor requirements, but performance is obviously insufficient filtering method, in line with a principle median filtering research-oriented image processing. Information processing for wireless sensor networks, research has important theoretical and practical significance for non-image information processing median filter method. According to the characteristics of resource-constrained sensor nodes processing, this paper analyzes the characteristics of the existing median filtering algorithm applied to the non-image processing, and pointed out that the median filtering for wireless sensor node data processing applications in noise suppression, the calculation speed and frequency selection capability shortcomings. Sensor nodes, efficient processing of the collected data, the paper proposed the pseudo median filtering of an improved method, by introducing a simple the pulsating noise detection link, the method still has enhanced the pseudo median filtering Ripple ability while the fast filtering characteristics; strengthen the pulsating noise suppression, further proposed a pseudo-median - median hybrid filtering method based on soft-switching strategy, the method combines the advantages of the median filter and the pseudo-median filter. Oriented frequency domain filtering applications, this paper, the one with negative weighting coefficient of the weighted repeat pseudo median filter design methods, and gives a simplified calculation algorithm, design according to the method of pseudo-median filter can be obtained better than the existing weighted the pseudo median filtering method selected frequency characteristics. Reference wireless sensor networks typical node configuration, this choice of the the MSP430F1611 microprocessor and IAR tools to build a test median filtering algorithm experiment platform. Based on the platform, the pulse wave signals measured data for the test sample to test the effectiveness of the improved median algorithm. Simulation and experimental tests show that median filtering this improved algorithm and design meet the needs of wireless sensor nodes signal processing, data processing applied to vibration, pulse wave, temperature, and other non-image information, and therefore has a broad application prospects .