Study On Wireless Sensor Network Node Location Technology And Kernel Scheduling Mechanism

Wireless Sensor Network(WSN)monitors the physical location or environmental conditions by a large number of sensor nodes.The position information and positioning technology of sensor nodes is a hot research topic in wireless sensor networks.Wireless Sensor Network Operating System(WSNOS)is the basic software environment of wireless sensor network,and the scheduling mechanism is the core techniques of operating system.So it has very important significance to study node localization algorithm and kernel scheduling mechanism.Since it does not require more hardware,the DV-HOP algorithm has the advantage of low cost and low energy consumption but has the disadvantage of lower localization precision.Therefore,in order to improve the positioning precision,this algorithm proposes a DV-HOP localization algorithm based on amended hops by using RSSI.Although Tiny OS has been widely used as a wireless sensor network operating system,its FCFS(First-Come-First-Served)scheduling method may cause a series of problems such as blocking the task and so on.Therefore,this paper propose a congestion avoidance strategy based on dynamic priority scheduling to optimize the system kernel.The main works are as follows:(1)A DV-HOP localization algorithm based on hops amended by using RSSI is proposed.In order to solve the problem in original algorithm by which the integer number of hops will reduce the positioning accuracy,this algorithm firstly calculated the distance that between the unknown nodes and anchor nodes by using the empirical formula of RSSI path loss model.Then the ratio of the theoretically distance to the communication radius was used as a new way of hops calculation.Finally,An correction model based on Gauss distribution function was established,In order to reduce the influence of average distance error to the precision in the original DV-HOP,the algorithm calculated the limit value of error function.By the experimental analysis,compared with others algorithm,the improved algorithm can reduce the positioning error effectively,and also has good stability.(2)A kind of dynamic priority scheduling based on congestion avoidance strategy is proposed.In order to improve the throughput and real-time performance of FIFO scheduling,this scheduling algorithm firstly modified the Kernel Source code of Scheduler,therefore the system will perform the task queue with high priority,this step can improve the real-time performance of TinyOS.Then based on the TCP congestion avoidance(additive increase and multiplicative decrease)algorithm,this proposed algorithm achieves the additive increasing and multiplicative decreasing of task queue by adding some variables and functions,and improves the network's throughput.By the experimental analysis,compared with FIFO scheduling policy,the improved scheduling mechanism improves the throughput and the real-time performance of TinyOS system effectively.