Research On Direction And Velocity Measurement Of Moving Item With UHF RFID

The content of this paper is to use RFID (Radio FrequencyIdentification) and Discrete-time Signal Processing technology formoving item to determine the direction and velocity.To determine the direction and velocity of moving item with RFIDtechnology has always been a hot application, which is very significant tothe fields like modern logistics, traffic detection, automatic security, etc.Almost all the major RFID vendors spend a lot of money and effort here.Many RFID reader manufacturers have products in the UHF band toachieve a similar function. In this paper, by summing up the previousspeed systems theory, implementation and detailed analysis of the UHFband RFID systems based on the EPC Global Agreement, presents acomputer-assisted, real-time to accurately determine the direction andvelocity of moving items solution. Compared with previous methodologywith RFID technology, this solution do not need a database, which canquickly adapt to complex and changing environment and get moreaccurate results to meet the practical needs of the industry. Additionally,this solution is easy to implement in engineering. Get information about moving item by the Communication Modulebetween computer and RFID reader. With the information and DataAnalysis Module, the system could determine the direction and velocityof the moving item, and then presents the final result on the screen byDisplay module. Compared with other methodology with RFIDtechnology, this paper does not require pre-established database of model,and solve the problem of calculation failure due to read failure caused bythe multi-path effect using the similarity of the same environment the twotags on the moving item encountered. This method can better adapt to thephysical environment, draw a rather precise velocity measurement resultswith a strong practical..In the Communications Module, this paperprovides an optimized protocol implementation to improve access to datarate and reliability. In the Data Analysis Module, this paper utilizeswavelet denoising, smoothing, interpolation, pattern extraction andrelated matching algorithms to estimate the direction and velocity of themoving item. In the wavelet denoising process, this paper presents animproved threshold method to remove white noise, effectively improvingSNR (signal to noise ratio). In model extraction and match correlation,this paper presents an improved matching algorithm, select the longestpiece of continuous signal as template and correlation match with theother signal to improve the matching accuracy. At last, verify the practicality and accuracy with the experiment withthe UHF RFID reader based on the Intel R1000.