This paper explores Mobile Crowd Sensing (MCS) within smart cities and the Internet of Things (IoT), focusing on enhancing data collection efficiency and quality in smart cities, intelligent transportation, and environmental monitoring. Addressing the challenges of uneven task distribution, low participant engagement, and subpar data quality in MCS, we propose a novel multi-task allocation scheme that leverages mobility prediction to optimize task allocation. By forecasting participants' movements, our optimization model aims to balance task completion rate, coverage, execution time, and worker distribution, thereby improving task allocation rationality and participant engagement. Our contribution lies in a novel task allocation mechanism that not only tackles current issues but also integrates mobility prediction, enhancing MCS data collection efficiency and quality. This work offers new insights into MCS system design and practical solutions for smart city and environmental monitoring applications, advancing MCS technology application.