Estimating Traffic Speed using Cellular Phone Data

Abstract

Traffic speed is an important parameter of traffic flow, particularly during morning or evening rush hours. It could be measured by using inductive loops or cameras, but this is restricted to critical crossroads or streets due to the cost of installations and management. In recent years, using vehicles with cellular phones as probes has become a hot topic to estimate traffic flow on roadways because it takes full advantage of existing cellular network systems to detect the movement of cellular phones. The challenge is to find an effective way to convert the rough positions of cellular phone bearers in cellular systems to exact routes in road network systems. This paper proposes a comprehensive approach that identifies regular commuters by analyzing their historical travel patterns, detects commuting routes using the algorithm of Mixed-Integer Linear Programming (MILP) with constraints, and thereafter estimates current traffic speeds. The approach was implemented using an Open Source GIS solution (Python Plus NetworkX and PostgreSQL) and tested with cellular phone data collected in the city of Haikou, China.