RFID Path Data Compression And Frequent Pattern Mining Based On IOV

RFID, radio frequency identification technology, utilizes radio frequency communicationto accomplish non-contact automatic identification, which can be fast, high effective,safeapplication for identification and storage of objects literacy in motion.One of the applicationsof RFID is telematics, that is, the electronic tag loaded on the vehicle is identified by RFIDradio frequency, which is to extract and analyze the static and dynamic properties informationof the vehicle.Besides, the electronic tag can also extract the data RFID path. The path datacan be compressed to reduce the space of storage data. Then frequency pattern mining usedon the RFID path data can understand the movement routes and moving trend of the vehicle,which can assist the relevant decision-making processes and ultimately design the facilitatetransportation planning and so on.In the aspect of the RFID path data compression, it is initial to use EPC code as the keyvalue to integrate route information from both the location and time dimensions due to thespatial and temporal characteristics of RFID data path.When compressing the positioninformation of RFID data path, an XML repeated distributionprime number coding MBPrimeis applied to RFID path coding. It improves the allocation strategy of the prime numbersbased on the PrimeLabel coding method. Therefore, the code of each path can be presentedwith the code of the last position in the same path.When compressing the whole routeinformation of RFID path data, the dictionary compression method is employed to improvethe threshold value determination, which is called the R-LZW compression algorithm.There are two methods proposed in the aspect of the frequency pattern mining of RFIDpath. The first one, which named PathMining depends on closed sequence mining algorithmCloSpan. Construct the lexicographic sequence tree LST as the prority of depth for pathdatabase.The LST can reflects the searching space of frequent path. In the process of structingLST,merge the equivalent subtree and the delete the non-closed sequences among theproduction of the candidate sequence sets. It can reduce the unnecessary research,be adaptedto mine the frequent patterns from RFID path data. The second method is mine the closedfrequent pattern through stream sliding window, proposing an ICR-Tree structure whichbased on the concept of inverse sequence. The new structure determine the path if thefrequent depends on the type of nodes. Finally, introduce two level hash index structure toexamine if the path is closed, and used the attenuation factor dynamically adjust he minimumsupport for optimizing algorithm, enable ICR-Stream algorithm would be more efficient inmining closed pattern from a sliding window.In this paper, the algorithm is verified by the simulation data,the simulation results showthat the MBPrime coding method can efficiently complete coding processing and R-LZW al- -gorithm can compress and store all the information of path. In the frequent pattern of path,PathMining algorithm has advantage in time,however,ICR_Stream algorithm has the highercoverage rate.