Optimizing Task Throughput for Large-scale Mobile Crowdsourcing in Smart Cities

Abstract

In mobile crowdsourcing, workers are financially motivated to perform as many self-selected tasks as possible to maximize their revenue. Unfortunately, the existing task scheduling approaches in mobile crowdsourcing fail to consider task execution duration and do not scale for massive tasks and large geographic areas. In this dissertation, we propose a novel framework, Turbo-GTS, in support of large-scale Geo-Task Scheduling (GTS), with the objective of identifying an optimal task assignment for each worker in order to maximize the total number of tasks that can be completed for an entire worker group while taking into account various spatial and temporal constraints, such as task execution duration, task expiration time, and worker/task geographic locations. Since the exact solution to the GTS problem is computationally intractable, we first propose two sub-optimal approaches (LCPF and NUD-IC) based on particle filtering and DBSCAN for the Single Worker Geo-Task Scheduling (SGTS) problem. We then extend our work to solve the Multi-Worker Geo-Task scheduling (MGTS) problem by proposing two space partitioning-based methods (QT-NNH and QT-NUD), which leverage the point-region quadtree to ensure workload balancing. We further propose WBT-NNH and WBT-NUD, which build on the algorithms QT-NNH and QT-NUD respectively, and provide more effective and dynamic workload balancing among all workers using the proposed Workload-balancing Bisection Tree (WBT) comparing to QT-NNH and QT-NUD. The effectiveness and efficiency of the six proposed approximate solutions are verified by our extensive experiments using both real and synthetic data. Compared with the state-of-the-art approaches, our proposed solutions are able to return a higher number of completed tasks for the worker group while reducing the computation cost by up to three orders of magnitude when coping with massive tasks distributed in large geographic areas. Last but not least, we present a web-based demo application for Turbo-GTS featuring the aforementioned four MGTS algorithms, which includes an interactive interface for users to load the current task/worker distributions and compare the task assignment of each worker returned by different algorithms in a real-time fashion. We also demonstrate the front-end interface of Turbo-GTS demo with several exploratory use cases in New York City.