群智感知系统中基于高效可信的任务分配关键技术研究

The past few years have witnessed the proliferation of smart mobile devices (such as smartphones, wearable devices, and etc.) in people’s daily lives. With the advent of emerging wireless communication technology (e.g. 4G) and more powerful processors, people can make use of the smart mobile devices to get the ubiquitous environment data from the real world. Further, people can use these data to support our daily lives and business work through analyzing the sensing data. Crowd sensing application has been regarding as a new paradigm that takes advantage of the pervasive smartphones to collect and analyze data. In a crowd sensing system, incentive mechanism and tasks assignment mechanism design play an important role, which can attract more users to achieve the task with a high quality. Auction has been widely used for resource allocation in wireless communications, thus it can provide enough incentive for the participants in the crowd sensing system. Unfortunately, there are many shortcomings in the existing crowd sensing studies. For instance, some critical factors (such as the heterogeneity of the sensing tasks, unreliable participants, competition between multi-crowdsourcer, and privacy preserving) have not been fully considered in the existing studies. Thus, our project focuses on the sensing tasks allocation by introducing the auction theory while taking strategy-proofness, profit of the crowdsourcer, and maximization of the social efficiency as the designing targets. In order to improve the performances of the proposed allocation mechanisms, our aim is to design a series of auction mechanisms for crowd sensing system that can satisfy varies demands of the different applications. Further, we choose to adopt the data set from the real-world to evaluate the performance of the proposed allocation mechanisms.

随着智能移动终端的功能越来越强大，人们可以方便地获取物理世界中的各类数据。群智感知正是在此基础之上发展起来的一个新的研究领域。其中，如何设计合理的激励机制和任务分配机制，以吸引到足够多的用户高质量地完成任务是决定群智感知系统成败的关键。拍卖技术可以为群智感知提供足够的激励。然而，现有研究仍存在诸多不足，未能充分考虑群智感知系统中存在的任务的异质性、用户的不可靠性、多数据消费者之间的竞争性和数据消费者的隐私保护需求等问题。为此，本课题将在综合考虑上述问题的基础上，围绕群智感知中的各类任务拍卖问题展开研究，分别以保证诚实性、实现数据消费者收益和社会效益的最大化、实现最大最小公平以及保护数据消费者隐私等为设计目标，采用拍卖理论与方法，研究设计满足不同应用需求的一整套群智感知任务分配机制和算法，并使用真实数据集进行仿真验证，以提高群智感知任务分配机制的综合性能。

本项目针对现有群智感知任务分配研究中存在的若干问题，例如缺乏合理有效的激励机制、为充分考虑用户的可靠性问题、无法保护参与者的隐私以及不支持多数据消费者模型等，提出了群智感知系统的任务分配模型研究、面向群智感知系统的异质任务分配问题研究、基于用户可靠性的群智感知任务分配机制研究、面向多数据消费者的最大最小公平任务分配机制研究、面向多数据消费者的诚实任务分配机制研究以及群智感知任务分配中的数据消费者隐私保护机制研究共6点研究内容。截至2019年12月底，针对所有研究内容均提出了与之相对应的激励机制及任务分配算法，相关算法在真实数据集进行验证；基于群智感知思想设计完成了一个城市路况的信息采集系统；为群智感知系统研究进一步的产业化打下了坚实的基础。在此基础上，项目组成员在国内外知名学术期刊会议共发表论文41篇(均标注项目编号)，其中SCI收录19篇，EI收录19篇，成果包括了CCF A类顶级会议或期刊6篇；B类顶级会议或期刊7篇；CCF C类会议期刊11篇；中科院JCR分区一区期刊3篇)，共申请发明专利5项，已授权1项；总体来说超额完成了既定计划。