Pareto-Based Bi-Objective Optimization for Multi-Task Assignment in Mobile Social Crowdsensing

Mobile Social Crowdsensing (MSCS) represents a new paradigm that enables users to perform tasks and access services through mobile smart devices and social networks. However, current task assignment frameworks primarily concentrate on individual or group-based task execution, which often results in high communication costs and suboptimal utilization of user’s collection capabilities. To address these challenges, we propose a novel method, called Pareto-Based Bi-Objective Optimization for Multi-Task Assignment (PBOTA). PBOTA begins with the construction of a social network model that effectively accommodates both individuals and groups based on real-world application scenarios. Following this, we design two distinct mechanisms to identify Pareto optimal solutions tailored for two types of tasks. For simple task assignments (STs), we develop a Pareto front construction mechanism that employs the weighted sum approach, integrating the two objective functions into a unified single-objective optimization problem. In contrast, for complex task assignments (CTs), we implement a Pareto optimal selection mechanism based on a backtracking algorithm. This mechanism identifies candidate user sets that satisfy the Karush-Kuhn-Tucker (KKT) conditions to facilitate dual-objective optimization while minimizing both platform utility and waste contributions. We conduct comprehensive experiments utilizing real datasets, which demonstrates PBOTA’s superior performance in comparison to existing methods.