Performance Evaluation of Quality of Service (QoS) by using Hybrid Algorithms in VANETs

VANETs are a kind of MANET in which each vehicle serves as a node and has a wireless router. The increasing number of car connections in the car communications system has heightened the demand for efficient solutions to urban traffic congestion problems in recent years. Recent scientific developments have revealed that many of the earlier types of systems focused on constructing inference engine systems based on a model of vehicular congestion that identifies and decreases congestions. However, VANET congestion control methods face several challenges, such as long delivery times, inefficient service use, insufficient bandwidth utilization, overhead connections, and so on. In this article, we proposed a model that combines these techniques (K-means and Scatter Search (SS)). These methods are used to improve network QoS by addressing the problem of congestion control. Using the K-means technique, we clustered the messages (size, validity time, emergency, etc.). After clustering the message, the Scatter Search (SS) method finds all feasible pathways between a source and destination node. These paths are decided by the sort of message generated by the K-means algorithm. As there are different paths from sender to the receiver, the SS algorithm sorts each route depending upon response message. The NS3.33 simulation is applied to assess the findings by providing appropriate simulations for the suggested network connection. There are three metrics (PDR, Overhead, and E2E Delay). The results showed an increase in the PDR of 92 % when the data rate is four packets. But decrease the PDR to 84 % when the data rate becomes 20 packets. The term of overhead recorded 20 MB when the number of vehicles is 50. While a overhead rises to 100 MB with increase the number of vehicles is increased to 250. In terms of E2E delay, the proposed system recorded (30 ms) in a 50-vehicle group of vehicles. Moreover, the number of end-to-end delays is increased to (400 ms) for 250 vehicles.