A Study Of Task Distribution Method For Mobile Crowd Sensing System

With the rapid development of information technology and sensor technology,wireless terminals such as smart phones are widely popularized.These devices are highly integrated with a lot of sensing elements and computation chips.They have powerful computing and sensing abilities which makes mobile crowd sensing becoming a hot research area.Distributing sensing task plays a key ro le in mobile crowd sensing system.How to design an efficient task distribution method in order to make the nodes receiving the task map faster and low-cost is an important issue worthy of study.To solve the problem that raster-map-based task distribution method has a large amount of redundant data and does not support progressive transmission in Zoom,a raster-vector mixed task distribution method framework for mobile crowd sensing system is proposed.The method can effectively reduce the data flow during distributing tasks.Besides,the method can also support progressive transmission.The main work is as follows:(1)Firstly,we analyze the mobile crowd sensing network.Mobile crowd sensing network structure and application is described in this paper.We also analyze the deficiencies of existing research results.Meanwhile,we also introduce several data structure used in the multi-scale expression of vector information and give a brief analysis of the characteristics.(2)Secondly,we introduced the raster-vector mixed task distribution method of mobile crowd sensing framework.We describe the method of defining sensing task and the way to define the sensing task map.The mobile node can interact with the server to achieve the task distribution and task receiving.Description of encoding rastervector mixed task information is given.According to the task map defined earlier,server encodes it into task information by raster-vector mixed process which is used to be distributed.Experimental results show that our method can save more data downloads compared to Zoom's pure raster way.(3)Finally,we describe the progressive transmission method for task information proposed in this paper.Illustrating the principle of progressive transmission,describing the task classifying method,block method and task street curve compression method.By using Douglas-Peucker algorithm,we can not only compress task information,but also use the results to generate incremental data for progressive transmission.After the mobile node enters the sensing area,they receive task information in a progressive manner.Experimental results show that by using Douglas-Peucker algorithm at a certain threshold value,the data can be compressed greatly with a small error percentage.Besides,experiments also show that under the conditions of downloading the same amount of data,by using progressive transmission method can avoid the need to receive the task information completely before decoding the whole task information compared to normal method,despite that it takes a little more data to record the positions of incremental information.