FTAM: A Fuzzy Traffic Adaptation Model for Wireless Mesh Networks

Wireless mesh networks (WMNs) are considered as the next step towards providing a high-bandwidth network over a specific coverage area. Because of their advantages over other wireless networks, WMNs are undergoing rapid progress and inspiring numerous multimedia applications such as video and audio real-time applications. These applications usually require time-bounded service and bandwidth guarantee. Therefore, there is a vital need to provide Quality of Service (QoS) support in order to assure better quality delivery. However, providing QoS support for real-time traffic in WMNs presents a number of significant technical challenges. In this paper, we focus on one of the most critical technical issues in QoS support, by proposing a novel QoS traffic adaptation model based on fuzzy logic theory, named FTAM, which is capable of supporting real-time traffic such as video and voice services. By monitoring the rate of change in queue length in addition to the current length of the queue, FTAM is able to provide a good measurement of the future queue state, and then to achieve the convenient traffic adaptation according to the network state. Index Terms—Wireless Mesh Networks; QoS; Traffic Adpta- tion; Fuzzy Logic; I. INTRODUCTION The last few years have witnessed a wealth of research ideas on Wireless Mesh Networks (WMNs) that are moving rapidly toward implemented standards. Although WMN research is a relatively new field it is gaining more popularity for various new applications. For instance, multimedia application that opens up for converged services and new purposes is quickly becoming a key focus area for wireless mesh communications (1). With the increase in both the bandwidth of wireless chan- nels and the computing power of mobile devices, it is expected that video and audio services will be offered over WMN in the future. However, enabling multimedia communications over such networks is remaining a challenging task for both academic and industrial communities. Video and audio real- time services typically require stringent bandwidth and delay guarantees. This makes the deployment of Quality of Service (QoS) mechanisms a vital need for the satisfaction of user's requirements. Real-time applications generate traffic at varying rates and usually require the network to be able to support such a changing rate. Therefore, providing QoS guarantees is crucial for supporting disparate services envisioned for future wireless mesh networks (2). Despite the efforts made to alleviate this issue, there still exist a number of barriers to the widespread deployment of real-time applications. The most prominent one is how to en- sure the traffic adaptation in the case of heavy congestion case. It is important to note that the existing solutions developed for wired networks can not be deployed directly within WMNs. Difficulties with these models lie in the fact that they are not adapted to different node states and resource variation, as in mesh environments the available bandwidth for each node varies with time since the medium is shared (3). In this paper, we introduce a novel QoS model for traffic adaptation based on fuzzy logic that is capable of support- ing real-time traffic such as video and voice services. A major factor behind using fuzzy logic theory to ensure the traffic adaptation, is its adequation to the uncertainty, the heterogeneity and the information incompleteness of WMN environment characterized by dynamic traffic changes. Our proposed model, build on both MAC and network layers, will bring about the benefits of the advances in the areas of artificial intelligence and wireless networking. The evaluation of the model performances will be studied under different traffic and network conditions. The balance between network performances and reliability when transmitting multimedia traffic is an important issue to consider too. The remaining of the paper is organized as follows. A brief description of Fuzzy Logic theory is provided in Section II. Section III gives a state of the art regarding QoS fuzzy models in WMN. Section IV presents our proposed model. In Section V, we discuss the performance evaluation of FTAM, while Section VI concludes the paper. II. FUZZY LOGIC