A Cognitive Approach for Performance Enhancement of Wireless Mesh Networks

Providing the required quality of service (QoS) for clients in a wireless mesh network (WMN) is a challenging task. This is due to the complex architecture of the WMN and unpredictable behavior of the wireless links and the intermediate nodes. This can be alleviated by introducing a cognitive mechanism within the WMN. In this paper, a WMN is developed, where the data traffic is forwarded in a cognitive manner via multiple paths. The aims are to maximize the client´s achievable performance (e.g., data rate, delay) and minimize the imposed congestion in the network. The proposed mechanism constantly perceives the network performance through feedback loops, learns and predicts the performance of the paths, and takes appropriate decisions. The decision includes a set of routing paths and their corresponding data rates. The resulted network is capable of adapting itself with the changes, whilst improving its performance. The proposed mechanism is simulated using OMNET discrete event simulator and subsequently validated.