移动群智感知中基于用户意愿的多任务分配模型
Multi-Task Allocation Based on User Willingness in Mobile Crowd Sensing

DOI: 10.12677/CSA.2021.117199, PP. 1941-1948

Keywords: 用户意愿，指针网络，多任务分配，移动深度学习
User Ambition, Pointer Network, Multi-Task Allocation, Mobile Deep Learning

Full-Text   Cite this paper   Add to My Lib

Abstract:

以群智感知平台为中心的任务分配方法中，平台收集用户实时位置、社交关系等信息并进行集中处理，将多个任务分配给用户，以实现感知时间最小化、感知成本最小化等优化目标。但传统方法没有着重考虑用户的执行意愿，与用户意愿相悖的分配方案和大量隐私信息的泄露不利于维护用户的长时参与水平。因此提出基于用户意愿的任务路径分配模型PtrNet-TA，在用户移动端通过本地历史感知数据训练一个seq2seq的指针网络模型，任务分配时将云端服务器提供的待执行感知任务集合作为输入传入本地指针网络模型从而输出符合用户意愿的任务路径分配给对应的用户。通过仿真实验表明，PtrNet-TA序列预测准确度优于分别采用LSTM模型与LSTM + Attention模型的基准方法。
In the platform-centric task allocation method, the MCS platform collects information such as real-time location, historical movement trajectory, and social relationship of the user for centralized processing, and assigns multiple tasks to users, so as to optimize targets such as minimizing perceived time and minimizing perceived cost. The conventional method does not focus on the user’s willingness to execute. The distribution scheme which is contrary to the user’s will and the disclosure of a large amount of private information are not conducive to maintaining the long-term participation level of users. Therefore, a task path allocation method PtrNet-TA based on the user’s intention is proposed, and a seq2seq pointer network model is trained on the user mobile terminal through local history sensing data, and the to-be-perceived task set provided by the cloud server is input as an input during task assignment. The local pointer network model is passed in to output a task path that matches the user’s wishes to the corresponding user. Simulation experiments show that PtrNet-TA is better than the benchmark method using LSTM model and LSTM + Attention model.

References

[1]	郭斌, 翟书颖, 於志文, 等. 群智大数据: 感知、优选与理解[J]. 大数据, 2017(5): 60-72.
[2]	Wang, J., Wang, F., Wang, Y., et al. (2018) HyTasker: Hybrid Task Allocation in Mobile Crowd Sensing. IEEE Transactions on Mobile Computing, 19, 598-611.
[3]	Nesa, N. and Banerjee, I. (2019) SensorRank: An Energy Efficient Sensor Activation Algorithm for Sensor Data Fusion in Wireless Networks. IEEE Internet of Things Journal, 6, 2532-2539. https://doi.org/10.1109/JIOT.2018.2871469
[4]	Wang, J.T., Wang, L.Y., Wang, Y.S., Zhang, D.Q. and Kong, L.H. (2018) Task Allocation in Mobile Crowd Sensing: State-of-the-Art and Future Opportunities. IEEE Internet of Things Journal, 5, 3747-3757. https://doi.org/10.1109/JIOT.2018.2864341
[5]	Estrada, R. (2017) A Crowd-Sensing Framework for Allocation of Time-Constrained and Location-Based Tasks. IEEE Transactions on Services Computing, 13, 769-785.
[6]	Xiao, M., Wu, J., Huang, L., Cheng, R. and Wang, Y. (2017) Online Task Assignment for Crowdsensing in Predictable Mo-bile Social Networks. IEEE Transactions on Mobile Computing, 16, 2306-2320. https://doi.org/10.1109/TMC.2016.2616473
[7]	Messaoud, R.B. and Ghamridoudane, Y. (2016) QEMSS: A Se-lection Scheme for Participatory Sensing Tasks. International Conference on Protocol Engineering (ICPE) and Interna-tional Conference on New Technologies of Distributed Systems (NTDS), Paris, 22-24 July 2015. https://doi.org/10.1109/NOTERE.2015.7293493
[8]	Wang, J., Wang, Y., Zhang, D., et al. (2018) Multi-Task Al-location in Mobile Crowd Sensing with Individual Task Quality Assurance. IEEE Transactions on Mobile Computing, 17, 2101-2113. https://doi.org/10.1109/TMC.2018.2793908
[9]	Hu, Y., Dai, G., Fan, J., et al. (2016) BlueAer: A Fine-Grained Urban PM2.5 3D Monitoring System Using Mobile Sensing. Proceedings of the 35th IEEE International Conference on Computer Communications, San Francisco, 10-14 April 2016, 1-9. https://doi.org/10.1109/INFOCOM.2016.7524479
[10]	Azzam, R., Mizouni, R., Otrok, H., et al. (2016) GRS: A Group-Based Recruitment System for Mobile Crowd Sensing. Journal of Network and Computer Applications, 72, 38-50. https://doi.org/10.1016/j.jnca.2016.06.015
[11]	Liu, Y., Guo, B., Wang, Y., et al. (2016) TaskMe: Mul-ti-Task Allocation in Mobile Crowd Sensing. Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing, Heidelberg, 12-16 September 2016, 403-414. https://doi.org/10.1145/2971648.2971709
[12]	Wang, L., et al. (2018) Mobile Crowd Sensing Task Optimal Allo-cation: A Mobility Pattern Matching Perspective. Frontiers of Computer Science, 12, 231-244. https://doi.org/10.1007/s11704-017-7024-6
[13]	Yu, H., Shen, Z., Miao, C., et al. (2013) A Reputation-Aware Decision-Making Approach for Improving the Efficiency of Crowdsourcing Systems. Adaptive Agents and Multi-Agents Systems, 13, 1315-1316.
[14]	Vinyals, O., Fortunato, M. and Jaitly, N. (2015) Pointer Networks. In: International Conference on Neural Information Processing Systems, MIT Press, Cambridge, 1-9.
[15]	Rashid, T., Samvelyan, M., De Witt, C.S., et al. (2018) QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning. Journal of Machine Learning Research, 21, 1-51.
[16]	Mnih, V., et al. (2015) Human-Level Control through Deep Reinforcement Learning. Nature, 518, 529-533. https://doi.org/10.1038/nature14236

Full-Text