Accurate Intelligent Map Matching Algorithms for Vehicle Positioning System M. Pashaian 1 , M. R. Mosavi 2 and M. Pashaian 3 1,2 Department of Electrical Engineering, Iran University of Science and Technology Narmak, Tehran 16846-13114, Iran Corresponding Author, 3 Department of Civil Engineering, Islamic Azad University, Science and Research Branch Tehran 14155-4933, Iran Abstract The purpose of a navigation system installed on a car is to help drivers in order to select an optimal path to reach the destination. In most of these systems, Global Positioning System (GPS) are used to determine vehicle position. There are number of error sources that undermine the quality of GPS measurements for car navigation systems. For this reason, technique like Map Matching (MM) is required to identify the road of a car is that moving on, with a high degree of confidence. MM in car navigation systems, has a task of determining the current position of vehicle on the city's map. In this paper, two new MM methods based on Fuzzy Logic (FL) and Neural Network (NN) are proposed to solve the matching problem in car navigation systems. For the experiments, a car navigation system is implemented with a low cost GPS receiver. The proposed fuzzy algorithm is easy to calculate. It requires little computation time without need to extra sensors and can find effectively the mobile exact position which moves on the road. Keywords: GPS, Map Matching, Fuzzy Logic, Neural Network. 1. Introduction Along with the development of Global Positioning System (GPS) technology nowadays, the positioning and navigation systems for vehicles have become the main applications in the intelligent transportation system, which brings great benefits in technology and economy, and get more and more attentions from the public. Therefore, high system performance is required urgently, especially the aspect of GPS positioning accuracy which is the key factor that affects the whole system accuracy [1]. As a matter of fact, the positioning error in GPS caused by various reasons is inevitable such as inherent errors from satellite ephemeris or GPS receiver, propagation delay through the ionosphere or any other complex unknown factors [2,3]. Existing methods applied in foreign Vehicle Navigation Systems (VNSs) have been proved effective in improving the GPS positioning accuracy by using of dead reckoning technique, differential GPS, the radio beacon or high-precision carrier phase receiver, etc [4,5]. As a result, the direct overlay of positional data obtained from GPS is not reconciled with a digital map. Therefore GPS data need to be corrected with various methods to match with a digital map. One of the important parts in the car navigation system is determining the position of the car on the map. The methods to find car's location can be classified: dead- reckoning that calculates car's location according to its direction and distance travelled, and radio navigation that pinpoints the location of a car in terms of absolute position through radio waves [6], and GPS. The GPS receiver output is then map matched to the road network in order to give the drivers information about their location on the map [7,8]. In VNS, Map Matching (MM) approach plays an important role. It is a method of using digital map data and GPS satellite signal to locate the vehicle on proper position related to digital map. MM is a technique that well performed in estimating the actual vehicle location from GPS position in use of the road information contained in the digital map according to the spatial relationship between the GPS position and the roads. Compared to the other methods in this field, there are no additional equipment costs and optimization in solving MM problems is obvious, the request for digital maps is much easier to be applied with an acceptable performance. However, MM methods usually cost a lot due to their dependence on external assist devices and the related technologies are also a bit complicated, which unfortunately limit their promotions in practical applications [9]. Fuzzy Logic (FL) is one technique that is an effective way to deal with qualitative terms, linguistic vagueness, and human intervention [10]. Neural Network (NN) is an information-processing device composed of highly interconnected nodes, the processing elements, that is inspired by the way biological nervous systems process information, such as the human brain [1]. FL and NN have IJCSI International Journal of Computer Science Issues, Vol. 9, Issue 2, No 3, March 2012 ISSN (Online): 1694-0814 www.IJCSI.org 114 Copyright (c) 2012 International Journal of Computer Science Issues. All Rights Reserved.