A multi-path cognitive resource management mechanism for QoS provisioning in wireless mesh networks

  • Authors:

  • Farshad Javadi;Abbas Jamalipour

  • Affiliations:

  • School of Electrical and Information Engineering, University of Sydney, Sydney, Australia 2006;School of Electrical and Information Engineering, University of Sydney, Sydney, Australia 2006

  • Venue:
  • Wireless Networks
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

In a Wireless Mesh Network (WMN), achieving acceptable Quality of Service (QoS) levels requires distributed control over network resources and subsequent awareness of the dynamically changing conditions of the WMN. In this paper, for facilitating such control, a cognitive mechanism is introduced, which facilitates cooperation and cognition among multiple Mesh Access Points and edge routers called Mesh Portals for routing client traffic via multiple paths. The aim of the cognition is to reasonably maximize the fulfillment of the clients from the achieved QoS (e.g., end-to-end delay and bandwidth). The cognitive process consists of three cycles. In the first cycle, the Perception Cycle, the current performance status of the WMN is continuously perceived through feedback loops. The perceived information is further processed and fed into the second cycle, the Learning Cycle, in order to understand the network conditions. This results in the prediction of the performance of the paths and estimation of the path delay for various load conditions. The third cycle, the Decision Cycle, is a game theoretic coalition formation algorithm, that results in path selection and data rate assignment. This algorithm is modeled as a cooperative game theory, which incorporates the Bilateral Shapley Value to find the best coalition from available paths, whereupon a bargaining game theory formulates the data rate assignment. Extensive simulations are performed for evaluating the proposed cognitive mechanism under various load conditions and results demonstrate the evident enhancement of the achieved end-to-end QoS of the clients and the network performance compared with non-cognitive scenarios, specifically in congested conditions.