Foggy-Park: A Dynamic Pricing and NSGA based Allocation Scheme for On-Street Parking System

In past decade parking and problems associated with it have attracted the researchers attention towards it. Some of thewell known problems associated in the path of making parking smart are optimal parking resources usage, guaranteed parking reservation, identification of available parking slots, efficient communication protocols. This paper proposes a scheme, namely, Foggy-Park which deals with dynamic pricing and allocation aspects of the smart on-street parking system. While allocating the available parking slots, Non-dominated Sorting Genetic Algorithm (NSGA) is used to address the interests of both the parkers and parking authorities. The parkers always desire to pay less parking fees. Whereas, the parking authorities want to generate high revenue by renting out parking slots. In order to compute dynamic prices for the available parking slots, Seattle city parking and its prices data-sets are used. The former one is used to train random forest model which is used predict occupancy. Whereas, the later one is used to form base prices. Foggy-Park scheme is implemented on different computing paradigms, such as, cloud, fog, and edge using the concept of Zero Trust Network Access (ZTNA). The scheme implemented on fog computing paradigms shows its worth over others in terms of less communication overhead. The obtained results prove that the proposed Foggy-Park scheme minimizes the average parking prices, maximizes the generated revenue, maximizes the accepted requests, and maximizes the occupancy fairness by around 4%, 23%, 6%, and 11.28% respectively.