1 CROSS CORRELATION AND CRITICAL PATH ANALYSIS BASED ROAD TRAFFIC ROUTING ALGORITHM Rapelang K. Kemoabe 1 , Ibo Ngebani 2 Abstract: In this paper we present a new dynamic road traffic routing algorithm, for enhanced clearing out of traffic from congested areas of the traffic network. The main objective of designing this algorithm is to implement a system that optimizes the rate of flow of traffic throughout the road network by minimizing traffic congestion rates. By cross – correlation analysis the system analyzes the relation between lanes, in particular the rate of flow of traffic between lanes. It is used to compute the Time Difference of Arrival (TDOA), which in this context estimates the amount of time a fleet of traffic takes to travel from one intersection to the next. These time estimates are then used together with traffic counts at each intersection, and traffic weights which depict traffic flow patterns between lanes, to compute link scores and path scores for each road link and path, respectively. These scores are then presented as a network model based on the concerned road part of the road network. After all this, the algorithm uses this network model to compute an optimized sequential opening of the traffic lights in within that particular area. This algorithm is a dynamic model that mimics the operation of a police officer who controls traffic flow at road intersections during bad traffic congestion, the practice which is common here in Botswana and other third world countries. Keywords: Road traffic, Routing Algorithm, Ant Colony Optimization (ACO), Ant Dispersion Routing (ADR), Critical Path Analysis, Poisson Probability Distribution, Cross Correlation Analysis. I. Introduction Road traffic congestion is proving to be a problem across the globe, and with motor vehicle numbers increasing significantly by the day. This phenomenon can only get worse if necessary measures are not taken. Traffic congestion can be caused by many factors ranging from vehicle breakdowns, bad roads, bad weather, faulty traffic control lights, high densities of vehicles on the roads and inefficient signaling methods, just to mention but a few. This problem has many implications in the economies and livelihoods of any country. A lot of effort has been put in to curb this problem, with the development of new road traffic routing algorithms. Ant Colony Optimization (ACO) techniques have been used widely in the design of these algorithms; with Alves et al (2010) proposing a dynamic algorithm based on ACO called Ant Dispersion Routing (ADR) Algorithm [1]. Krömer et al (2011) presented a dynamic traffic routing approach that focused on the enhancement of routing in problematic areas of the road network like, accident scenes, road works and traffic jams, also based on ACO. The algorithm called Ant Colony Inspired Algorithm for Adaptive Traffic Routing uses a probability threshold to achieve sensitive and dynamic routing in a multi-path road network [2]. Kelly et al (2007) presented an inter – lane interactive algorithm called Decentralized Car Traffic Control using Message Propagation Optimized with a Genetic Algorithm [3]Shashikiran et al (2011) proposed a dynamic traffic routing protocol, for road traffic management during peak hours based on Krushkal’s and Dijkstra’s algorithms. In this protocol Krushkal’s algorithm is used to implement traffic routing and guidance part, while Dijkstra’s algorithm is used to perform the optimal path discovery procedure based on the information availed by the vehicles’ Dynamic Vehicles Navigation System (DVNS) [4]. BeeJamA Algorithm proposed by Wedde et al (2009) is a multi – agent bottom – up self - adaptive traffic routing protocol that is based on the behavior of the honey bee colony as they endure on their food searching endeavors [5]. In Botswana and in other developing countries police officers are tasked with the responsibility to control the traffic flow at the congested segments of the city road network during peak hours. This system has thus far proved to be working and highly efficient as compared to currently used conventional system which uses pre-programmed traffic lights, which are unable to adapt to the variations in traffic density due to their fixed timing signaling. We therefore; here present a novel approach in traffic routing based on Critical Path Analysis (CPA) and Cross – Correlation Analysis (CCA). This traffic routing technique is a model that mimics the operation of the police officer system. CPA is generally known as network analysis (NA). It can essentially be employed in any multi-task complex project to ensure that the full project is completed in the most minimum time and as efficiently as possible [6]. In this research, we seek to design a system that will ensure the maximum flow rate of traffic within a city road network. Cross-correlation analysis is a mathematical technique that can be used as a measure of relation between any given time series [7]. It can also be used in the estimation of the Time Difference of Arrival (TDOA), which is widely used in applications like multi-channel synchronous data acquisition, real time data processing and positioning [7, 8, 9]. The rest of the paper is organized as follows: The System Model is presented in Section II; Road Traffic Control Routing Algorithm Based On CPA is presented in Section III; Results and Analysis in Section IV and finally Conclusions in Section V.