Incentive Mechanism for Mobile Crowd Sensing with Assumed Bid Cost Reverse Auction

TL;DR

This paper introduces a novel reverse auction mechanism for mobile crowd sensing that reduces resource waste and improves system efficiency by allowing users to bid before data collection and dynamically recruit new participants.

Contribution

The paper proposes RA-ABC and RA-ABCDR, innovative auction models that outperform existing methods by enhancing user retention and reducing costs in mobile crowd sensing.

Findings

RA-ABCDR achieves 54.6% higher user retention.

RA-ABCDR reduces auction costs by 22.2%.

Dynamic recruitment improves system stability and fairness.

Abstract

Mobile Crowd Sensing (MCS) is the mechanism wherein people can contribute in data collection process using their own mobile devices which have sensing capabilities. Incentives are rewards that individuals get in exchange for data they submit. Reverse Auction Bidding (RAB) is a framework that allows users to place bids for selling the data they collected. Task providers can select users to buy data from by looking at bids. Using the RAB framework, MCS system can be optimized for better user utility, task provider utility and platform utility. In this paper, we propose a novel approach called Reverse Auction with Assumed Bid Cost (RA-ABC) which allows users to place a bid in the system before collecting data. We opine that performing the tasks only after winning helps in reducing resource consumption instead of performing the tasks before bidding. User Return on Investment (ROI) is calculated with which they decide to further participate or not by either increasing or decreasing their bids. We also propose an extension of RA-ABC with dynamic recruitment (RA-ABCDR) in which we allow new users to join the system at any time during bidding rounds. Simulation results demonstrate that RA-ABC and RA-ABCDR outperform the widely used Tullock Optimal Prize Function, with RA-ABCDR achieving up to 54.6\% higher user retention and reducing auction cost by 22.2\%, thereby ensuring more efficient and sustainable system performance. Extensive simulations confirm that dynamic user recruitment significantly enhances performance across stability, fairness, and cost-efficiency metrics.