Research On Resource Selection Strategy In Mobile Group Intelligence Perception Platform

In recent years,with the development of wireless network,mobile swarm intelligence perception network has become a hot topic in network science research.Most of the research focuses on how to use better incentive mechanism to motivate users,how to screen better participants,how to improve data quality,etc.However,there is a lack of research on resource selection strategy in group intelligence perception platform.This paper studies this issue,and the specific work is as follows:This paper constructs a multi-task,multi-user and multi-choice mobile crowd sensing perception platform.In this platform,according to the different characteristics of users and tasks,naive bayesian classification algorithm is used to classify them in order to improve the accuracy of resource selection.For uncertain tasks and users,user-centered task selection strategy and task-centered user selection strategy are designed respectively.Among these two resource selection strategies,this paper designs three resource allocation algorithms:pervasive,optimal and greedy.In the pervasive algorithm,resources are selected according to the order of integration,and reservation mechanism is added to the optimization algorithm.In the greedy algorithm,resources selection is strengthened.In addition,this paper also designs an integral incentive mechanism based on bidding linkage.In this mechanism,rthe creditworthiness of publisher and user is maintained by using the integral strategy.The publisher accumulates the integral through publishing tasks,and the user accumulates the integral through completing tasks.In this integration strategy,the task price is linked with integral through the game between publisher and user on task price.It guarantees the competitiveness of the platform,breaks the monopoly of task price,and finally encourages users to participate in more tasks through incentives.This paper proves that the utility of the platform is non-negative through theoretical deduction.Through simulation experiments,it is verified that the task-centered user selection strategy guarantees the greatest degree of resource selection in greedy mode and user-centered task selection strategy guarantees the greatest degree of resource selection in optimal mode with the increase of tasks.With the increase of users,both resource selection strategies are optimal mode resources.The selection is the strongest.It also verifies that with the increase of users and tasks,the three resource allocation modes have different resource selection strengths in the platform.When the number of users is fixed,the more tasks the greedy mode accomplishes;when the number of tasks is fixed,the more tasks the user optimizes the mode accomplishes;and in the pervasive mode,the amount of tasks the user accomplishes fluctuates between the two modes.