大规模社会空间移动众包可信招募过程IoT
A Trustworthy Recruitment Process for Spatial Mobile Crowdsourcing in Large-scale Social IoT
摘要：
空间移动众包（SMCS）可以通过利用社交物联网（SIoT）的功能来执行空间任务。通常，在SMCS中，任务请求者旨在招募IoT设备的子集并委托它们前往任务位置。然而，由于物联网网络及其多样化设备（例如，多个品牌、不同的沟通渠道等），招募合适的设备/工人正在成为一项具有挑战性的任务。为此，在本文中，我们开发了一个招聘过程SMCS平台使用自动SIoT服务发现，以选择值得信赖的工人满足请求者的要求。该方法主要包括两个阶段：1）工作者过滤阶段，旨在使用应用于SIoT关系图的Louvain社区发现算法（CD）将工作者的搜索空间减少到潜在可信候选者的子集。接下来，2）选择过程阶段，其使用线性规划（ILP）来确定所选设备/工作器的最终集合。ILP最大限度地提高了工作人员的效率指标，包括技能/规格水平、招聘成本和招聘的IoT设备的可信度。选定的实验分析所提出的CD-ILP算法使用真实世界的数据集的性能，并显示其优越性，在提供一个有效的招聘策略相比，现有的随机算法。
英文摘要：
Spatial Mobile Crowdsourcing (SMCS) can be leveraged by exploiting the capabilities of the Social Internet-of-Things (SIoT) to execute spatial tasks. Typically, in SMCS, a task requester aims to recruit a subset of IoT devices and commission them to travel to the task location. However, because of the exponential increase of IoT networks and their diversified devices (e.g., multiple brands, different communication channels, etc.), recruiting the appropriate devices/workers is becoming a challenging task. To this end, in this paper, we develop a recruitment process for SMCS platforms using automated SIoT service discovery to select trustworthy workers satisfying the requester requirements. The method we purpose includes mainly two stages: 1) a worker filtering stage, aiming at reducing the workers' search space to a subset of potential trustworthy candidates using the Louvain community detection algorithm (CD) applied to SIoT relation graphs. Next, 2) a selection process stage that uses an Integer Linear Program (ILP) to determine the final set of selected devices/workers. The ILP maximizes a worker efficiency metric incorporating the skills/specs level, recruitment cost, and trustworthiness level of the recruited IoT devices. Selected experiments analyze the performance of the proposed CD-ILP algorithm using a real-world dataset and show its superiority in providing an effective recruitment strategy compared to an existing stochastic algorithm.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
