| In this dissertation, first, a closed-loop congestion control strategy is developed. This strategy is a dynamic and distributed congestion control strategy that detects the congestion by measuring the channel usage level. Then, the congestion is controlled by tuning the transmission range and rate that considerably impact on the channel saturation. Tuning the transmission range and rate in VANets is an NP-hard problem due to the high complexity of determining the proper values for these parameters in vehicular networks. Considering the benefits of Tabu search algorithm and its adaptability with the problem, a multi-objective Tabu search algorithm is used for tuning transmission range and rate in reasonable time. In the proposed algorithm, the delay and jitter are minimized as the objective functions of multi-objective Tabu Search algorithm.;Second, two open-loop congestion control strategies are proposed that prevent the congestion occurrence in the channels using the prioritizing and scheduling the messages. These strategies define the priority for each message by considering the content of messages (i.e. types of the messages for example emergency, beacon, and service messages), size of messages, and state of the networks (e.g. velocity, direction, usefulness, distance and validity metrics). The scheduling of the messages is conducted based on the defined priorities. In addition, as the second scheduling technique, a Tabu Search algorithm is employed to schedule the control and service channel queues in a reasonable time. For this purpose, the delay and jitter of messages delivery are minimized.;Finally, a localized and centralized strategy is proposed that uses RSUs set at intersections for detecting and controlling the congestion. These strategy clusters all the messages that transferred between the vehicles stopped before the red traffic light using Machine Learning algorithms. In this strategy, a K-means learning algorithm is used for clustering the messages based on their features (e.g. size of messages, validity of messages, and type of messages, and so on). The communication parameters including the transmission range and rate, contention window size, and Arbitration Inter-Frame Spacing (AIFS) are determined for each messages cluter based on the minimized delivery delay. Then, the determined communication parameters are sent to the vehicles by RSUs, and the vehicles operate based on these parameters for transferring the messages.;The performances of three proposed strategies were evaluated by simulating the highway and urban scenarios in NS2 and SUMO simulators. Comparisons were also made between the results obtained from the proposed strategies and the common used congestion control strategies. The results reveal that using the proposed congestion control strategies, the throughput, packet loss ratio and delay are significantly improved as compared to the other strategies. Therefore, applications of the proposed strategies help improve the performance, safety, and reliability of VANets. (Abstract shortened by ProQuest.). |
