摘要

随着智慧城市建设步伐的加速，环境感知需求日益增长并呈现复杂化趋势，移动群智感知，特别是车联网群智感知（Vehicular Crowdsensing，VCS）的出现给复杂环境下的城市感知带来了更精确有效的解决方案。根据任务空间分布的不同，VCS系统中存在不同的感知任务类型，其中位置依赖型任务因其时空分布的独立性被广泛应用于交通流检测和道路健康监测等场景，促进该类型任务的有效执行对于实现当前的城市感知需求具有重要意义。然而对于VCS系统中的车辆用户而言，参与感知会消耗其自身资源，并且存在隐私泄露的风险，它们往往不愿意主动参与，因此迫切需要研究有效的激励机制。另外，针对位置依赖型任务，VCS存在机会式和参与式两种感知模式，然而现有激励机制的研究在这两种感知模式下都面临着一些亟待克服的挑战。对于机会式感知而言，一方面用户隐私安全和数据安全得不到有效的保障，另一方面如何在有限的成本内激励用户稳定的提供高质量数据也是研究的难点所在。对于参与式感知而言，现有研究难以适应多任务并发的场景，在该场景下，现有激励机制中的任务分配方案在感知效率和系统效益方面均有待提升。

针对上述问题，本文提出一种面向机会式感知的安全保护和质量保障型激励机制以及一种多任务并发的参与式感知激励机制。本文的主要研究及创新点如下：

（1）针对机会式感知，本文创新性地融合区块链技术和面向数据质量的中标参与者选择算法，一方面保障了用户的隐私安全和数据安全，另一方面从确保感知数据来源的安全可靠和依据信誉对感知用户进行筛选两种途径共同提高数据质量，相比现有基线方法，能实现可控激励成本范围内数据质量62.28%的提升。另外，本文对现有基于反向拍卖的激励机制的流程进行了改进，在中标参与者选择和奖励分配之间加入数据质量评估和信誉更新，使得用户为获取更多奖励更倾向于持续稳定地提供高质量数据，仿真结果中也通过累积奖励的对比验证了这一结论。

（2）针对参与式感知，本文提出对于具有聚类特性的任务用任务组合代替单个任务进行批量分配，并设计了一种自适应任务组合算法，通过自适应地找到最优的聚类数目实现最优的任务组合效果，从而大大提高了多任务并发场景下的感知效率和聚类有效性。另外，本文首次提出一个两阶段的任务分配方案，在参与者选择阶段优化平台利益，在任务执行顺序规划阶段优化参与者利益，以达到系统整体利益最优的目的。最后的仿真结果表明，本文提出的多任务并发的参与式感知激励机制在感知效率和系统效益方面相比现有方案均实现有效提升。

ABSTRACT

With the acceleration of the construction of smart cities, the demand for environmental sensing is increasing and becoming more complicated. The emergence of Mobile Crowdsensing, especially Vehicular Crowdsensing (VCS), has brought a more accurate and effective solution to urban sensing in complex environments. According to different spatial distribution of tasks, there are different types of sensing tasks in the VCS system. In these types, location-dependent task is widely used in traffic flow detection and road health monitoring due to its independent spatial-temporal distribution. Promoting the effective execution of these tasks is of great significance for realizing the demand of current city sensing. However, for the vehicle users in VCS system, participating in VCS will consume their own resources and may have the risk of privacy disclosure. They are often unwilling to actively participate. Therefore, effective incentive mechanisms are urgently needed to be studied. In addition, for location-dependent tasks, VCS has two sensing modes: opportunistic sensing and participatory sensing. However, the existing researches on incentive mechanism face some challenges to be overcome in these two modes. For the incentive mechanism of opportunistic sensing, on the one hand, user privacy security and data security cannot be effectively guaranteed, on the other hand, how to incentive users to provide high-quality data stably within the limited cost is also the difficulty of the research. For the incentive mechanism of participatory sensing, the existing research is difficult to adapt to the multi-task concurrent scenario. In this scenario, the task allocation schemes in the existing works need to be improved in terms of sensing efficiency and system benefit.

 

In view of the above problems, this thesis proposes a secure and privacy protection and data quality assurance incentive mechanism for opportunistic sensing and a multi-task concurrent incentive mechanism for participatory sensing. The main research and innovation points of this thesis are as follows:

 

For opportunity sensing, this thesis innovatively integrates blockchain technology and data quality-oriented winning bidders selection algorithm. On the one hand, it ensures user privacy and data security. On the other hand, it improves data quality through two ways: ensuring the safety and reliability of sensing data sources, and filtering users based on reputation. Compared with the existing baseline, the data quality can be improved by 62.28% within the range of controllable incentive cost. In addition, this thesis improves the process of existing incentive mechanisms based on reverse auction by adding data quality assessment and reputation update between winning bidder selection and reward distribution. It allows users more inclined to consistently provide high-quality data in order to get more rewards. This conclusion is also verified by comparison of cumulative rewards in the simulation results.

 

For participatory sensing, this thesis proposes to use task combination instead of single task for batch assignment of tasks with clustering characteristics. Then this thesis designs an adaptive task combination algorithm. The optimal task combination effect is realized by finding the optimal cluster number adaptively, which greatly improves the sensing efficiency and cluster validity in multi-task concurrent scenario. In addition, a two-stage task allocation scheme is proposed for the first time in this thesis, which optimizes the benefit of platform at the participant selection stage and the benefit of participants at the task conduction sequence planning stage, so as to optimize the overall benefits of the system. The final simulation results show that the multi-task concurrent incentive mechanism for participatory sensing proposed in this thesis can effectively improve the sensing efficiency and system benefit compared with existing schemes.

[1] Zhang K, Ni J, Yang K, et al. Security and privacy in smart city applications: Challenges and solutions[J]. IEEE Communications Magazine, 2017, 55(1): 122-129.
[2] Singh B, Lobiyal D K. An energy–efficient adaptive clustering algorithm with load balancing for wireless sensor network[J]. International Journal of Sensor Networks, 2012, 12(1): 37-52.
[3] Guo W, Zhu W, Yu Z, et al. A survey of task allocation: contrastive perspectives from wireless sensor networks and mobile crowdsensing[J]. IEEE Access, 2019, 7: 78406-78420.
[4] Bellavista P, Cardone G, Corradi A, et al. Convergence of MANET and WSN in IoT urban scenarios[J]. IEEE Sensors Journal, 2013, 13(10): 3558-3567.
[5] Capponi A, Fiandrino C, Kantarci B, et al. A survey on mobile crowdsensing systems: Challenges, solutions, and opportunities[J]. IEEE communications surveys & tutorials, 2019, 21(3): 2419-2465.
[6] Hu J, Yang K, Wang K, et al. A blockchain-based reward mechanism for mobile crowdsensing[J]. IEEE Transactions on Computational Social Systems, 2020, 7(1): 178-191.
[7] Zhang C, Zhu L, Xu C, et al. PRVB: Achieving Privacy-Preserving and Reliable Vehicular Crowdsensing via Blockchain Oracle[J]. IEEE Transactions on Vehicular Technology, 2020, 70(1): 831-843.
[8] 李静林, 袁泉, 杨放春. 车联网群智感知与服务[J]. 中兴通讯技术, 2015, 21(6): 6-9.
[9] Gao G, Xiao M, Wu J, et al. Truthful incentive mechanism for nondeterministic crowdsensing with vehicles[J]. IEEE Transactions on Mobile Computing, 2018, 17(12): 2982-2997.
[10] Zhang X, Yang Z, Sun W, et al. Incentives for mobile crowd sensing: A survey[J]. IEEE Communications Surveys & Tutorials, 2015, 18(1): 54-67.
[11] Peng D, Wu F, Chen G. Data quality guided incentive mechanism design for crowdsensing[J]. IEEE transactions on mobile computing, 2017, 17(2): 307-319.
[12] Tao X, Song W. Location-dependent task allocation for mobile crowdsensing with clustering effect[J]. IEEE Internet of Things Journal, 2018, 6(1): 1029-1045.
[13] Wang J, Wang Y, Zhang D, et al. Multi-task allocation in mobile crowd sensing with individual task quality assurance[J]. IEEE Transactions on Mobile Computing, 2018, 17(9): 2101-2113.
[14] Gong W, Zhang B, Li C. Task assignment in mobile crowdsensing: Present and future directions[J]. IEEE network, 2018, 32(4): 100-107.
[15] 杨朔. 移动群智感知中的数据质量分析和基于质量的用户管理[D]. 上海: 上海交通大学, 2019.
[16] Restuccia F, Ferraro P, Silvestri S, et al. IncentMe: Effective mechanism design to stimulate crowdsensing participants with uncertain mobility[J]. IEEE Transactions on Mobile Computing, 2018, 18(7): 1571-1584.
[17] Li M, Weng J, Yang A, et al. Crowdbc: A blockchain-based decentralized framework for crowdsourcing[J]. IEEE Transactions on Parallel and Distributed Systems, 2018, 30(6): 1251-1266.
[18] Wang J, Li M, He Y, et al. A blockchain based privacy-preserving incentive mechanism in crowdsensing applications[J]. IEEE Access, 2018, 6: 17545-17556.
[19] Ma Z, Huang W, Bi W, et al. A master-slave blockchain paradigm and application in digital rights management[J]. China Communications, 2018, 15(8): 174-188.
[20] Dorri A, Steger M, Kanhere S S, et al. Blockchain: A distributed solution to automotive security and privacy[J]. IEEE Communications Magazine, 2017, 55(12): 119-125.
[21] An J, Cheng J, Gui X, et al. A lightweight blockchain-based model for data quality assessment in crowdsensing[J]. IEEE Transactions on Computational Social Systems, 2020, 7(1): 84-97.
[22] Yang S, Wu F, Tang S, et al. On designing data quality-aware truth estimation and surplus sharing method for mobile crowdsensing[J]. IEEE Journal on Selected Areas in Communications, 2017, 35(4): 832-847.
[23] Yang F, Lu J L, Zhu Y, et al. Heterogeneous task allocation in participatory sensing[C]. 2015 IEEE Global Communications Conference (GLOBECOM). IEEE, 2015.
[24] Gong W, Zhang B, Li C. Location-based online task scheduling in mobile crowdsensing[C]. GLOBECOM 2017-2017 IEEE Global Communications Conference. IEEE, 2017.
[25] Li X, Zhang X. Multi-task allocation under time constraints in mobile crowdsensing[J]. IEEE Transactions on Mobile Computing, 2019, 20(4): 1494-1510.
[26] Wang L, Yu Z, Zhang D, et al. Heterogeneous multi-task assignment in mobile crowdsensing using spatiotemporal correlation[J]. IEEE Transactions on Mobile Computing, 2018, 18(1): 84-97.
[27] Zhao Y, Zheng K, Li Y, et al. Destination-aware task assignment in spatial crowdsourcing: A worker decomposition approach[J]. IEEE Transactions on Knowledge and Data Engineering, 2019, 32(12): 2336-2350.
[28] Wang J, Wang Y, Zhang D, et al. PSAllocator: Multi-task allocation for participatory sensing with sensing capability constraints[C]. Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social Computing. 2017.
[29] Xu Q, Zheng R. When data acquisition meets data analytics: A distributed active learning framework for optimal budgeted mobile crowdsensing[C]. IEEE INFOCOM 2017-IEEE Conference on Computer Communications. IEEE, 2017.
[30] Zhang X, Yang Z, Liu Y. Vehicle-based bi-objective crowdsourcing[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(10): 3420-3428.
[31] Wang J, Wang Y, Zhang D, et al. Energy saving techniques in mobile crowd sensing: Current state and future opportunities[J]. IEEE Communications Magazine, 2018, 56(5): 164-169.
[32] 李静林, 刘志晗, 杨放春. 车联网体系结构及其关键技术[J]. 北京邮电大学学报, 2014, 37(6): 95.
[33] Yang F, Wang S, Li J, et al. An overview of internet of vehicles[J]. China communications, 2014, 11(10): 1-15.
[34] 陈荟慧, 郭斌, 於志文. 移动群智感知应用[J]. 中兴通讯技术, 2014 (1): 35-37.
[35] Magerkurth C, Cheok A D, Mandryk R L, et al. Pervasive games: bringing computer entertainment back to the real world[J]. Computers in Entertainment (CIE), 2005, 3(3): 4-4.
[36] Avouris N M, Yiannoutsou N. A review of mobile location-based games for learning across physical and virtual spaces[J]. J. Univers. Comput. Sci., 2012, 18(15): 2120-2142.
[37] Fudenberg D, Tirole J. Game theory[M]. MIT press, 1991.
[38] 杨熳. 基于区块链技术的会计模式浅探[J]. 新会计, 2017 (9): 57-58.
[39] 李文森, 王少杰, 伍旭川, 等. 数字货币可以履行货币职能吗?[J]. 新理财: 公司理财, 2017 (6): 25-28.
[40] Xiong J, Xie H, Liu B, et al. Cooperative Caching Services on High-Speed Train by Reverse Auction[J]. IEEE Transactions on Vehicular Technology, 2021, 70(9): 9437-9449.
[41] Kang J, Yu R, Huang X, et al. Blockchain for secure and efficient data sharing in vehicular edge computing and networks[J]. IEEE Internet of Things Journal, 2018, 6(3): 4660-4670.
[42] 吴卫明. 数字金融法律实务与风险防控[M]. 北京: 中国法制出版社, 2018.
[43] 李春燕. 基于区块链技术的车载自组织网络路况预警方案研究[D]. 北京: 北京交通大学, 2019.
[44] 苗世清, 高岳林. 求解 0/1 背包问题的离散差分进化算法[J]. 小型微型计算机系统, 2009 (9): 1828-1830.
[45] Montazeri M, Kiani R, Rastkhadiv S S. A new approach to the restart genetic algorithm to solve zero-one knapsack problem[C]. 2017 IEEE 4th International Conference on Knowledge-Based Engineering and Innovation (KBEI). IEEE, 2017.
[46] Myerson R B. Optimal auction design[J]. Mathematics of operations research, 1981, 6(1): 58-73.
[47] Yuan J, Zheng Y, Xie X, et al. Driving with knowledge from the physical world[C]. Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining. 2011.
[48] Yuan J, Zheng Y, Zhang C, et al. T-drive: driving directions based on taxi trajectories[C]. Proceedings of the 18th SIGSPATIAL International conference on advances in geographic information systems. 2010.
[49] Li Q, Cao H, Wang S, et al. A reputation-based multi-user task selection incentive mechanism for crowdsensing[J]. IEEE Access, 2020, 8: 74887-74900.
[50] Bhattacharjee S, Ghosh N, Shah V K, et al. QnQ: Quality and Quantity Based Unified Approach for Secure and Trustworthy Mobile Crowdsensing[J]. IEEE Transactions on Mobile Computing, 2018, 19(1): 200-216.
[51] Yu H, Yang Y, Zhang H, et al. Reputation-Based Reverse Combination Auction Incentive Method to Encourage Vehicles to Participate in the VCS System[J]. IEEE Transactions on Network Science and Engineering, 2021, 8(3): 2469-2481.
[52] Sun H, Yi J, Xu Y, et al. Crack Monitoring for Hot-Spot Areas Under Time-Varying Load Condition Based on FCM Clustering Algorithm[J]. IEEE Access, 2019, 7: 118850-118856.
[53] Bezdek J C, Moshtaghi M, Runkler T, et al. The generalized c index for internal fuzzy cluster validity[J]. IEEE Transactions on Fuzzy Systems, 2016, 24(6): 1500-1512.
[54] Yelmewad P, Kumar A, Talawar B. MMAS on GPU for large TSP instances[C]. 2019 10th international conference on computing, communication and networking technologies (ICCCNT). IEEE, 2019.