Blockchain Based Parking Pricing Prediction for Intelligent Transportation System

Abstract

Increasing growth of vehicles leads to traveling issues like traffic congestion, accidents, and parking management. To handle aforementioned issues, Intelligent Transportation System (ITS) is continuously working. Nowadays, searching for parking is a tedious and time consuming task. Also, static parking pricing leads to traffic congestion and air pollution problem in parking areas.

Furthermore, ITS mainly focuses on roadside traffic and road issues, while parking problems are generally not considered. However, parking a vehicle consumes a major time of traveling trip. Hence, more research is required on parking management and pricing to reach up to the optimal value of parking price. Several existing techniques and articles exist in literature for parking management and its price prediction. Though it has not been explored fully and comprises various issues like data theft, failure and others. Hence, we have proposed a Machine Learning (ML) and BC-based integrated approach to predict parking availability and parking pricing rates based on dynamic pricing mechanism, where "cost is low with low demand and cost is high with high demand". Here, BC protects transaction history to enhance security and data privacy. So, we have designed an approach that consists neural network-based parking fees prediction model integrated with BC-based data security. The model is analyzed using several metrics for example Mean Absolute Error (MAE), Mean Squared Error (MSE), Root-Mean-Square Error (RMSE) and others.