Decentralized Optimal Parking Lot Allocation via Dynamic Parking Fee.

This paper proposes a novel decentralized optimal parking lot allocation algorithm by using dynamic parking fee. Information about drivers and parking lots used to be collected by a system manager in smart parking systems, and the allocation is determined centrally. However, the computational load increases exponentially as the target area expands. Therefore, a decentralized system is needed that can flexibly respond to the expansion of the area size, in which the allocation is determined by communicating the results of the calculations performed by each participant. In building a decentralized system, it is desirable for the system to be fair and efficient. Therefore, we propose a decentralized parking system that satisfies this property by using matching theory, a field that studies how people and things should be matched in a market. We propose a unique matching method that takes into account the difference between drivers as consumers and parking service providers. We also propose a reallocation method based on dynamic parking fee to reduce the number of unallocated drivers.