Routing and Channel Allocation in Rural Wireless Mesh Networks

IEEE 802.11 Wi-Fi equipment based wireless mesh networks have recently been proposed as an inexpensive approach to connect far-flung rural areas. Such networks are built using high-gain directional antenna that can establish long-distance point-point links. In recent work, a new MAC protocol named 2...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Dutta, P., Jaiswal, S., Rastogi, R.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:IEEE 802.11 Wi-Fi equipment based wireless mesh networks have recently been proposed as an inexpensive approach to connect far-flung rural areas. Such networks are built using high-gain directional antenna that can establish long-distance point-point links. In recent work, a new MAC protocol named 2P has been proposed that is suited for the interference pattern within such a network. However, the 2P protocol requires the underlying graph (for each 802.11 channel) to be bi-partite. Under the assumption that 2P is the MAC protocol used in the mesh network, we make the following contributions in this paper. Given K non-interfering 802.11 channels, we propose a simple cut-based algorithm to compute K bi-partite sub-graphs (on each of which the 2P protocol can be run separately). We establish the class of graphs that can thus be completely covered by K bipartite subgraphs. For the remaining set of graphs, we look into the "price" of routing all end-to-end demands over only the bipartite subgraphs. We analytically establish what fraction of the max flow of the original mesh-graph can be routed over the bipartite subgraphs. Finally we look into the problem of mismatch between the load on a link (as computed by max flow) and its effective capacity under a given channel allocation. We propose heuristics to cluster links with similar loads into the same bipartite graphs (channels) and through comprehensive numerical simulations show that our heuristics come very close to the best possible flow.
ISSN:0743-166X
2641-9874
DOI:10.1109/INFCOM.2007.76