A tarp and ensemble classifier-based real-time congestion control approach for V2V communication
摘要
A vehicular Ad-hoc Network is an advanced concept for establishing inter-vehicular communication in which moving vehicles can exchange significant information like direction, location, speed, traffic congestion, etc. Network distributed organizing by self-communication mobile network created by moving vehicles, distinguished by very high mobility dependent on vehicle speed and degrees of movement freedom. In VANET, information exchange is the primary goal of distributing data from one node to the possible nodes in the network. But information is frequently duplicated, and the network becomes overloaded with packets, leading to packet drops and congestion. Lack of accuracy in determining important parameters like location, speed, direction, etc., leads to the selection of improper routes in the network. We proposed the Traffic Analysis-based Route Planning (TARP) mechanism to overcome these drawbacks. The proposed TARP applies an ensemble classifier, enhancing traffic flow by minimizing congestion. The route planner in the proposed system is very effective in determining alternate routes for drivers. To measure the proposed TARP efficiency an evaluation assessment is done with the existing error model-based adaptive rate control, combined power and data-rate adaptation, and linear message rate integration control (LIMERIC) mechanisms. The comparison assessment is conducted on the aspect of end-to-end delay, traffic rates, packet delivery ratio and routing overhead. On all the evaluation metrics, the proposed TARP performance is very effective than the others. As a result, the proposed TARP has achieved an accuracy of 99.3%, the maximum of the existing mechanisms.