Wireless mesh networks: performance and reliability studies

点击翻译按钮获取中文摘要

Wireless mesh networks (WMNs) consist of stationary wireless routers interconnected by wireless links with a small fraction acting as gateways to the Internet. The key challenge facing WMNs is supporting high throughput (in addition to security and privacy). The maximum available bandwidth degrades rapidly with increasing distance between the source and destination in the mesh network; this degradation is caused due to interference from neighboring hops and paths. The capacity of WMNs can be increased using multi-radio nodes with multiple (orthogonal) channels; also, with careful adaptation of transmission power of wireless routers, one can minimize interferences and thus increase capacity.In this project we intend to (a) develop new "generalized" models and tools for the design of WMNs with the primary goal of minimizing the cost of WMNs while satisfying all the physical and performance requirements; the uniqueness of our proposal lies in the fact that rather than treating each of the components (e.g., multi-channel, power control, cost, QoS, etc.) separately, as has been done so far in the literature, we will address the above issues using a unified approach by exploiting the intimate relationship among them; and (b) develop new adaptive management schemes to provide QoS support and assurance for different classes of users. The basic novel idea behind these schemes is to react to anomalies (e.g., failure and congestion) by reconfiguring (e.g., new channel assignment, new topology) the WMN while minimizing, ideally avoiding, the impact on the ongoing traffic.Using simulations platforms to evaluate and experiment with the models/protocols/algorithms, we are planning to develop, may produce inaccurate and/or misleading results (e.g., simpler models/heuristics are used to quantify interferences in simulation platforms). We will use the equipment (requested in this application) to evaluate and experiment with the proposed solutions. This will allow getting accurate and "real" results which may lead us to tune and improve the performance of our solutions.