A Survey of Task Allocation: Contrastive Perspectives From Wireless Sensor Networks and Mobile Crowdsensing

Wenzhong Guo, Weiping Zhu, Zhiyong Yu, Jiangtao Wang, Bin GuoIEEE Access 2019 · OpenReview

Abstract

Wireless sensor networks (WSNs) and mobile crowdsensing (MCS) are two important paradigms in urban dynamic sensing. In both sensing paradigms, task allocation is a significant problem, which may affect the completion quality of sensing tasks. In this paper, we give a survey of task allocation in WSNs and MCS from the contrastive perspectives in terms of data quality and sensing cost, which help to better understand related objectives and strategies. We first analyze the different characteristics of two sensing paradigms, which may lead to difference in task allocation issues or strategies. Then, we present some common issues in task allocation with objectives in data quality and sensing cost. Furthermore, we provide reviews of unique task allocation issues in MCS according to its new characteristics. Finally, we identify some potential opportunities for the future research.

Introduction: The Evolution of Urban Sensing

As cities evolve into "smart" environments, the demand for large-scale urban sensing—monitoring air quality, traffic, and noise—has grown exponentially. Historically, this was achieved through Wireless Sensor Networks (WSNs), which rely on dedicated, static hardware nodes. However, the emergence of Mobile Crowdsensing (MCS) has introduced a paradigm shift, leveraging the ubiquitous sensors in consumer mobile devices (smartphones, wearables) to collect data opportunistically.

The architecture of Mobile Crowdsensing (MCS) involves service requestors, a central platform, and mobile users.

The core challenge in both paradigms is task allocation: determining which sensing node (a fixed sensor or a mobile user) should perform a specific task. While WSNs are primarily constrained by hardware limitations like battery life, MCS introduces "the human factor," where mobility, privacy, and incentives become critical constraints.

Comparative Framework and Methodology

This survey categorizes task allocation research by contrasting the automated, deterministic nature of WSNs with the collaborative, opportunistic nature of MCS. The authors identify a taxonomy of issues that are either shared across both paradigms or unique to the human-centric MCS model.

Taxonomy of task allocation issues, contrasting WSN and MCS characteristics and identifying common vs. unique challenges.

Common Issues: Data Quality and Sensing Cost

Both WSNs and MCS must address two fundamental objectives:

  1. Data Quality: This involves ensuring adequate coverage and fault tolerance. In WSNs, coverage is often deterministic (e.g., a fixed radius), whereas in MCS, it is probabilistic and dynamic, depending on user movement.
  2. Sensing Cost: For WSNs, this focuses on energy conservation to extend network life. In MCS, sensing cost translates to user retention; excessive battery drain leads to users uninstalling sensing applications.

Unique MCS Challenges

The transition to human-centric sensing introduces several unique hurdles, including Incentive Mechanisms (motivating participants), Location Privacy, and Trustworthiness (filtering malicious data). Unlike programmed nodes, human participants require motivation and protection to remain active in the network.

Key Insights and Future Directions

The survey highlights that while MCS can adopt certain WSN techniques—such as "Primary/Backup" (P/B) copy methods for reliability—they must be augmented to handle human unpredictability. WSN strategies typically favor centralized scheduling, whereas MCS strategies rely on greedy selection based on predicted mobility and auction-based mechanisms.

Looking forward, the authors identify several promising avenues:

  • Hybrid Sensing: Integrating WSNs for baseline sensing with MCS for high-density urban areas.
  • Learning-Assisted Allocation: Utilizing Machine Learning to predict user "skill levels" and task completion likelihood.
  • Cluster-Based MCS: Borrowing the "Cluster Head" concept from WSNs to allow local data aggregation among mobile users, reducing communication energy costs.

Summary updated

Abstract

Wireless sensor networks (WSNs) and mobile crowdsensing (MCS) are two important paradigms in urban dynamic sensing. In both sensing paradigms, task allocation is a significant problem, which may affect the completion quality of sensing tasks. In this paper, we give a survey of task allocation in WSNs and MCS from the contrastive perspectives in terms of data quality and sensing cost, which help to better understand related objectives and strategies. We first analyze the different characteristics of two sensing paradigms, which may lead to difference in task allocation issues or strategies. Then, we present some common issues in task allocation with objectives in data quality and sensing cost. Furthermore, we provide reviews of unique task allocation issues in MCS according to its new characteristics. Finally, we identify some potential opportunities for the future research.
BibTeX
@inproceedings{guo2019survey,
  author = {Wenzhong Guo and Weiping Zhu and Zhiyong Yu and Jiangtao Wang and Bin Guo},
  title = {{A Survey of Task Allocation: Contrastive Perspectives From Wireless Sensor Networks and Mobile Crowdsensing}},
  booktitle = {IEEE Access},
  year = {2019},
  url = {https://openreview.net/forum?id=I5jEaUq88Y},
}
Figure