Research On Wear Leveling Of Data Storage In Sensor Node Based On Greedy Strategy

While working wireless sensor network generate massive data,which records lots of monitored area information in all aspects?ALL the information will be passed from the root to the routing node via data link and the network topology.And finally,monitoring personnel can get these useful information by the means of satellite and etc? In recent years,wireless sensor technology has been suffering more and more attention from both domestic and overseas,which result in a gradual mature in technology?Monitoring and Acquisition of various types of experimental data in the field is the fu ndamental task of WSN.WSN is a new network lines, the main core of its network is to sup port all kinds of physical information monitoring area, so however its monitoring of enviro nmental issues on the topology and transforms,what users most concern need are an effecti ve data transmission and storage, so as to ensure follow-up studies have efficient, reliable, r eal-time data.So how to get efficient, reliable, real-time data has become an important expl oration WSN problems.In this paper, the basic storage components in wireless sensor network NAND FLASH memory storage mode that is in the block level as the unit first erase in read and write operation process,Thus causing the problem that the physical block's unbalanced erased number of times to reduce the memory service life, A greedy algorithm is proposed to partition the address mapping wear leveling,When reading and writing data,first to find the logical address,Then the greedy selection is carried out according to the wear and tear coefficient,select a new physical block for reading and writing,the old physical block for hot and cold data migration,enter wait erase.Compared to other algorithms,the greedy algorithm can be given more quickly,getting a better solution.The greedy algorithm tries to solve the problem by gradual design and operation.It can be further optimized by the optimization of the system.Whenever a greedy judgment is made, the problem is subdivided into a relatively small problem,and then to a certain extent, the overall solution is also a set of high quality solutions under certain conditions.The NAND FLASH memory is simulated by the software method,to verify the design of the greedy partitioning algorithm for memory wear leveling and optimization.Compare this algorithm with the flash translation layer(FTL),the memory footprint of the algorithm is 512 K,system overhead is greatly reduced.In the comparison of threshold selection, the overall consideration of wear leveling and data writing efficiency is considered,in this paper, a threshold value is 50,which is chosen as the ideal threshold of the algorithm,so the data write efficiency of NAND FLASH memory is 66.59%,the number of erase times of the physical block is also more balanced.