基于压缩感知的移动群智感知任务分发机制

针对移动群智感知任务中区域全覆盖感知成本过高问题,提出基于压缩感知的移动群智感知任务分发(CS-TD)机制。首先提出了感知任务整体成本模型,该模型综合考虑了参与感知任务的节点个数、节点的感知次数与数据上传次数;然后基于成本模型,分析感知节点的日常移动轨迹,结合压缩感知数据采集技术,提出了一种基于感知节点轨迹的压缩感知采样方法;其次通过区域全覆盖最少节点(RCLN)算法,选出最佳节点集合,对节点进行任务分配,利用压缩感知技术恢复节点数据;最后在多次感知任务的迭代中对感知节点的可信程度进行评定,保证任务方案的最优性。对CS-TD分发模型进行多次实验验证,与已有的Crowd Tasker算法相比,CS-TD算法平均成本降低了30%以上。CS-TD模型能有效降低感知节点的消耗,能在全覆盖感知任务中降低整体感知成本。

Mobile crowdsensing task distribution mechanism based on compressed sensing

Since the cost of mobile crowdsensing in full coverage of area is excessively high, a Compressive Sensing-based mobile crowdsensing Task Distribution (CS-TD) mechanism was proposed. Firstly, an overall cost model of perceived task was proposed. In this model, the number of nodes participating in a perceived task, the number of nodes perceived and data uploaded were comprehensively considered. Then based on cost model, the daily movement trajectory of sensory node was analyzed, by combining with the compressed sensing data acquisition technology, a compressed sensing sampling method based on perceived node trajectory was proposed. Secondly, the optimal node set was selected by the Region Covers Least Nodes (RCLN) algorithm, the tasks were assigned to the nodes, and then the compressed sensing technology was used to recover node data. Finally, the trustworthiness of perceived node was evaluated in iteration of multiple perceived tasks to ensure the optimality of task plan. The CS-TD distribution model was tested several times. Compared with the existing CrowdTasker algorithm, the average cost of CS-TD algorithm is reduced by more than 30%. CS-TD model can effectively reduce consumption of sensing node and reduce overall perceived cost in full coverage sensing task.