Social-Network-Assisted Recruiting MCS Workers Strategy With Non-Fixed Reward

Mobile crowdsensing uses the mobility of workers and terminals with embedded sensors to complete complex perception tasks.Large-scale perception tasks require sufficient workers to provide high-quality data.However,in reality,the newly deployed mobile crowdsensing platform faces a pool of workers.Cold start problem caused by small size.Existing studies have used social network communication tasks and recruited workers to randomly generate task propagation probability or acceptance probability according to a certain probability distribution,and set the recruitment reward to be a fixed constant,which means that it cannot be applied to reality and efficient use of budget.In order to solve the above problems,this thesis studies the quantification model of the influence between users in social networks,and proposes a worker recruitment algorithm under non-fixed remuneration.Aiming at the recruitment of workers under nonfixed remuneration,this thesis proposes a two-stage solution to ensure the efficiency and correctness of the calculation.At the stage of sufficient budget,by estimating the difference in movement trajectories between workers and the range of influence of workers,users with larger trajectory differences and larger influence ranges are first selected to be placed in the seed set;when the recruitment remuneration of the seed set is estimated,it is close to the budget threshold.When entering the second stage,the incremental ratio of coverage to recruitment compensation after candidate workers join the seed set is estimated through simulation,and the greedy algorithm is used in each iteration to select the candidate worker with the largest incremental ratio to join the seed set.A series of experiments are carried out to evaluate the effect of the proposed model and the performance of the algorithm.The algorithm and the baseline algorithm are run on multiple real social network data sets,and the recruitment effect is judged by the coverage rate and running time of the algorithm.Experimental and analysis results show that the spatio-temporal coverage rate under the recruitment strategy under nonfixed remuneration is increased by 2.4%-13.7%compared with the baseline algorithm,while ensuring that the running time is within the tolerance range.