Wireless Mesh Networks


 
 
Project
Overview

 
 
BWN-Lab
Personnel

 
 
Project
Description

 
 
Publications

 
 
Testbed

 
 
BWN-Lab
Research Projects 

 
You are visitor

since 11/08/2004
Overview

Wireless mesh networks (WMNs) are undergoing rapid progress and inspiring numerous deployments. They are intended to deliver wireless services for a large variety of applications in personal, local, campus, and metropolitan areas. WMNs are anticipated to fundamentally resolve the limitations and to significantly improve the performance of wireless LANs, PANs, and MANs. They will greatly impact the development of wireless-fidelity (Wi-Fi), worldwide inter-operability for microwave access (WiMAX), Ultra Wide Band (UWB), wireless sensor networks and the newly emerging area of cognitive radios.

Mesh connectivity significantly enhances network performance, such as fault tolerance, load balancing, throughput, protocol efficiency; and dramatically reduces cost. Most of existing standard wireless networks do not have these capabilities. For example, standard Wi-Fi and WiMAX present typical point to- multipoint communication architectures. Networking protocols for mobile ad hoc networks (MANETs) and wireless sensor networks have considered multipoint-to-multipoint connectivity; however, their design have been focused on either issues of frequent topology changes due to high mobility, or a power efficiency mechanism. Protocols considering these constraints may be insufficient to reduce cost, enhance functionality, and improve performance for WMNs. Several challenges at each layer of the protocol stack need to be addressed to realize practical WMNs, such as, end-to-end and hop-based fairness and efficient flow management, choice of the transmission channels, formation of optimal routing paths in presence of high user traffic, amongst others. Recent experimental results have pointed out the impact of physical layer multi-path fading and co-channel interference as the key factors influencing packet delivery in a WMN. In addition, in a multi-channel environment, there exists significant power spectral overlap among channels used by MRs, leading to adjacent channel interference. Thus, it is of critical importance to investigate cross-layer approaches that take into account the underlying channel characteristic in the design of end-to-end protocols. We design such a cross-layer routing protocol in this project that jointly selects the transmission channel, the transmission rate, routing path while analytically estimating the end-to-end delay.

With the growing commercial deployments of WMNs and other WiFi networks, the 2.4 GHz ISM band is getting saturated. In order to relieve this congestion, future research in this area may target spectrum resource management functionalities. The mesh routers comprising a WMN may identify portions of vacant spectrum in altogether different frequency bands, e.g. at 700 MHz that is currently unused by the TV station that it is licensed to. Thus, making optimum use of the scarce spectrum resource in a decentralized manner marks the next stage of evolution in WMN research. We classify such networks as Cognitive WMNs and integrate principles from machine learning, algorithms, graph theory, physical layer signal processing into the classical networking approach to solve the research challenges.