SCSC2003 Abstract S3896
Predictive Quantization-based Filtering in Distributed Traffic Simulation
Predictive Quantization-based Filtering in Distributed Traffic Simulation
Submitting Author: Dr. JongKeun Lee
Abstract:
This paper proposed predictive quantization method that uses average speed-based dynamic quantum size to effectively reduce communication overhead that DEVS/HLA-based distributed microscopic traffic simulation has. Average speed-based dynamic quantum size monitors dynamically changing road simulation at fixed time unit and set vehicles¡¯ average speed to dynamic quantum size. Then it generates individual quantum size of factors such as the distance from the leading vehicle, speed and acceleration that decide microscopic traffic models¡¯ behavior by using dynamic quantum Size. Therefore, the suggested methodology has an advantage in that it effectively reduces communication overhead among federates of distributed microscopic traffic and output events among microscopic models in federates. This paper verified validity of the suggested method by doing distributed modeling the 32km block from East Seoul to Hobup of JOONGBOO highway in Korea and analyzing the simulation
Back to SCSC2003 Abstracts