Throughput optimization in DSRC for collision avoidance

Abstract

As a standard way of moving vehicles’ communication, VANETs (Vehicular Networks) have been characterized with enormous ability to increase the effectiveness of traffic and enhance the safety of cars on roads. Connected vehicle technology aims at dealing with important road transportation issues that concern environment and safety. Vehicular network density varies depending on the traffic load which can be high in the city, or low in suburban areas. The research provided a realistic analysis of the DSRC performance for varying vehicle densities over physical areas for urban and highway traffic using data packet response times and throughput as parameters. The Packet Delivery Ratio, the ratio of successfully received packets to the total number of packets sent, helps examine the hidden terminal effect on the communication using multi-lane traffic simulations for highway and urban scenarios. Computer simulations were done for 20, 40, 80 and 150 communication nodes for each traffic scenario; urban and highway to observe the effect of varying vehicle densities on the vehicular network communication performance. DSRC performance was studied for the traffic while analyzing the changes in the control parameters. We then proposed a control algorithm that secured data packets to address transmission errors resulting from the possible hidden node problem by improving packet throughput and controlling transmission delay. The algorithm minimizes packet losses resulting from possible collisions by giving an improved number of processes completed per unit time (throughput) for an increasing packet rate and effectively controlling the delay. Traffic simulation was done using a combination of SUMO (Simulation of Urban Mobility), OMNeT++ and VEINS for varying vehicle densities.