An Aggregate Parameter for Congestion Detection in VANETs

Vehicular ad hoc networks (VANETs), which are a class of Mobile ad hoc networks, have recently been developed as a standard means of communication among moving vehicles. Congestion is one of the key issues which can restrict the Network performance, and it is no exception for VANETs. Congestion control mechanisms include three phases: congestion detection, congestion notification and rate adjustment. Congestion detection is the first stage of congestion control mechanisms, and it is also the content what this paper focuses on. Because the available resources are limited, and the network topology and node density change over time, applying conventional congestion detection protocols to VANETs can be problematic. To address this challenge, this paper proposes an aggregate parameter based on weights for congestion detection, which can monitor the network performance from the four aspects: Messages Delivered Ratio (MDR), Average Delay (AD), Throughput and Overhead Ratio (OR), so as to achieve the purpose of the congestion detection. In addition, we introduce the method of normalizing the above four parameters, which make the aggregate parameter more accurate. Finally, simulation results show that the aggregate parameter is correct and effective.