Due to increasingly severe mobile cellular network congestion, especially during peak hours in the urban area, dynamic smart data pricing mechanisms have been proposed for service providers to shift users??? data consumption from peak hours to off-peak periods. Time-dependent pricing, as a major type of existing smart data pricing mechanism, fails to utilize location information and may perform poorly with spatially heterogeneous mobile traffic. In this article, we carry out a trace-driven analysis on the motivation and benefit of location-dependent pricing, based on a large-scale cellular network dataset including 9000 base stations and 3,500,000 subscribers. Our trace-driven analysis finds that location is another important factor that affects the consumption of mobile data, and the spatially heterogeneous feature should be considered in the design of pricing strategies. Through quantitative analysis, we reveal that even intuitively using location information in data pricing can significantly enhance system performance, reducing the peak-to-average ratio of traffic consumption by more than 15 percent in half of the base stations compared to pure time-dependent pricing.
展开▼
