Implementation of shortest remaining processing time based schedulers on a 32 bit serial based processing platform

Abstract

Vehicle-actuated traffic light controllers react to changes in traffic density throughout the day in order to optimize traffic flow effectively using traffic light control algorithms. Microscopic simulators are typically used to determine the performance of a traffic control scheme/algorithm by studying their effect on the relevant traffic parameters such as average waiting time, average travel time etc. However, the only method for real time evaluation of the control algorithms on physical platforms is to use pseudo-random number generation as input traffic data. In this paper we propose the use of a 32-bit serial based processing platform for the real time evaluation of traffic control schemes that utilizes accurate traffic data generated from a microscopic traffic simulator. This integrated platform can be used to evaluate traffic control algorithms more accurately and realistically since the algorithms are evaluated using traffic simulation data instead of random numbers.