MAPark: A Multi-Agent Auction-Based Parking System in Internet of Things

Abstract

By exploiting the ubiquitous nature of Internet of Things (IoT) technology in smart cities, we propose a Multi-Agent Auction-based Parking (MAPark) system for drivers as well as public and private Parking Facility Providers (PFP). In order to improve parking space allocation, the new idea in this paper is to incorporate parking auction with highly influential factors such as type of customer and parking, drivers’ and PFP’s preferences. MAPark leverages real-time demand-based pricing of parking spots, thus charging demand-based price to drivers while providing higher revenue to PFPs. The main contribution of this paper is the introduction of parking auction through sophisticated software agents that are capable of quickly and seamlessly performing tasks such as parking lookup, negotiation, pricing, and reservation. Simulation results show that the proposed scalable and interactive agents of MAPark enables efficient processing of a vast amount of data, providing cost savings in terms of time and fuel for drivers looking for parking, ultimately lowering traffic congestion, and improving usability of public and private parking facilities.