Sensing-gain constrained participant selection mechanism for mobile crowdsensing
- PDF / 2,387,619 Bytes
- 15 Pages / 595.224 x 790.955 pts Page_size
- 82 Downloads / 235 Views
ORIGINAL ARTICLE
Sensing-gain constrained participant selection mechanism for mobile crowdsensing Dan Tao1,2
· Ruipeng Gao3 · Hongbin Sun1
Received: 11 June 2020 / Accepted: 30 September 2020 © Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Participant selection is a fundamental challenge to perform sensing tasks with adequate data quality in various mobile crowdsensing (MCS) applications. In this paper, we explore participant selection mechanisms with sensing-gain constraints in MCS. First, we propose a novel quality-aware participant reputation model with active factors. Second, since user density differs in various applications, we propose two kinds of sensing-gain constrained participant selection mechanisms with both sufficient and insufficient user resources. Particularly, in the case of sufficient user resources, we formulate the sensing-gain objective on recruit cost and participant scale under constraints on data quality and task coverage, and propose a Multi-Stage Decision mechanism via Greedy strategy (MSD-G); in the case of insufficient user resources, we formulate the sensinggain objective on data quality, abstract it as a 0-1 knapsack problem, and devise a Sensing-Gain Constrained Dynamic Programming (SGC-DP) mechanism. Extensive simulations over a real-world dataset have verified that our participant reputation model with active factors can distinguish high-quality participants with different active levels, and our MSD-G and SGC-DP algorithms can effectively select suitable participants with ideal recruit budget and guaranteed data quality. Keywords Mobile crowdsensing · Data quality · Reputation model · Participant selection
1 Introduction Mobile crowdsensing (MCS) is a new paradigm of applications that utilizes ubiquitous mobile devices to collect and share the sensing data on surrounding environment over a large geographical region [1, 2]. Compared with traditional static sensor networks, there are many distinct advantages in MCS, including its lightweight deployment, low-cost maintenance, and flexible mobility. However, the data quality in MCS is affected by both subjective factors (e.g., meritorious or malicious behaviors) and objective factors (e.g., professional skills, device performances, and environment conditions) during data collection, user reputation
Dan Tao
[email protected] 1
School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing, 100044, China
2
The State Key Laboratory of Integrated Services Networks, Xidian University, Xi’an, 710071, China
3
School of Software Engineering, Beijing Jiaotong University, Beijing, 100044, China
and participant selection [3]. For example, continuous lowquality data may impede the service credibility of a sensing platform. In order to improve the data quality in MCS, existing approaches [4–6, 20] mainly focus on incentive mechanisms to recruit mobile users for collecting high-quality data in a specific sensing task. Jurca et al. [5] proposed a reputation model as the incentive mechanism, a
Data Loading...
