群智感知中基于时空关联性的用户激励机制

为在群智感知系统中实现有效的用户激励，提出基于显性与隐性时空关联的两种用户激励算法。将显性时空关联的用户激励问题转化为集合覆盖问题并利用贪心算法对其进行求解，同时结合显性时空关联算法和马尔科夫模型求解隐性时空关联的用户激励问题。在仿真数据和真实数据集上的实验结果表明，与传统最小化花费算法、最大化覆盖算法和最小化花费覆盖数比值算法相比，显性时空关联算法和隐性时空关联算法有效解决了感知任务完成率低且花费高的问题，能在实现用户激励的情况下最大化社会收益。

User Incentive Mechanism Based on Spatial-Temporal Correlation for Crowd Sensing

To realize effective user incentive in crowd sensing systems,this paper proposes two user incentive algorithms based on dominant and recessive spatial-temporal characteristics.The user incentive problem of dominant spatial-temporal correlation is transformed into a set coverage problem,which is solved by using the greedy algorithm.Then the dominant spatial-temporal correlation algorithm is combined with the Markov model to solve the user incentive problem of recessive spatial-temporal correlation.Experimental results on simulated data and real datasets show that compared with the traditional Minimum Cost(MC)algorithm,Maximum Overlap(MO)algorithm,and Minimum Cost Overlap(MCO)algorithm,the dominant and recessive spatial-temporal correlation algorithms can realize user incentive with the social benefits maximized and solve the problem of low completion rate and high cost of sensing tasks.