Map matching based on Cell-ID localization for mobile phone users

Abstract Map matching is a process of aligning a sequence of location estimates to a sequence of road segments in a road network to reduce the noisiness of the location estimates. Most existing map matching methods are designed based on GPS localization, which has many limitations (e.g. unstable urban operations and power hungry) and not suitable for mobile phone users. In this paper, we propose a map matching method based on Cell-ID localization for mobile phone users. Cell-ID localization is stable and energy efficient, but the location estimates are highly inaccurate, making the existing methods ineffective. For this problem, the proposed method firstly handles the inaccurate location estimates from Cell-ID localization through a series of preprocessing steps, and then uses a HMM (Hidden Markov Model) to align a sequence of location estimates to a sequence of road segments. The experimental results based on a real-world dataset collected in an urban environment have demonstrated the effectiveness of the proposed approach.