Elsevier

Computer Communications

Volume 31, Issue 8, 25 May 2008, Pages 1475-1483
Computer Communications

On the nominal capacity of multi-radio multi-channel wireless mesh networks

https://doi.org/10.1016/j.comcom.2008.01.056Get rights and content

Abstract

Wireless mesh networking is an emerging technology with a wide range of applications, from community to metropolitan area networking. Mesh networks consist of wireless routers and access points that make up the network backbone which is connected to the wired infrastructure. Client devices are located at the edge. The capacity of a wireless mesh network is an important performance indicator as well as design parameter. It depends on various factors such as network size and topology, expected traffic patterns, number of interfaces per node, number of channels, routing and channel assignment etc. In this paper, we present an analytical framework for estimating the capacity of wireless mesh networks, based on the concept of collision domains. The capacity of grid-oriented mesh networks is investigated for various scenarios to study the impact of different network parameters.

Introduction

Wireless mesh networks (WMNs) have generated a lot of interest in recent years, both in the industry as well as the research community. These networks can potentially be used in a wide range of applications such as last-mile access, enterprise networks, transportation systems, video surveillance systems [1]. A common feature of these WMN use cases is the presence of a wireless backbone connected to the fixed infrastructure via one or more wired links. This is the most commonly used mesh architecture, also referred to as infrastructure or backbone WMN. It consists of a set of mesh routers and access points connected by wireless links, with one or more of the nodes directly connected to the global Internet. The latter are referred to as gateways. Mesh routers and access points are typically fixed and hence, such WMNs have a fairly stable topology. In the following, unless mentioned otherwise, the term WMN refers to backbone mesh networks.

The design of a WMN for any application scenario is challenging because there are several interlinked factors that impact the performance of the network. Some of them are network-related, such as network architecture (size, density, topology etc.), traffic and node mobility patterns and so on. Other factors are radio related, for example, operating frequency band, channel bandwidth, number of available channels, number of radios per node etc. Different metrics can be used to quantify the performance of mesh networks. These include capacity, throughput, goodput, delay etc. Among these, capacity is without doubt one of the most important indicators. In this paper, a framework for estimating the capacity of multi-radio, multi-channel WMNs is described.

Section snippets

Related work

Capacity analysis of multi-hop wireless networks has proven to be an interesting and challenging research topic which is reflected in the wide body of related literature [2], [3], [4], [5], [6], [7]. In [3], lower and upper bounds of network capacity were determined. This work also provided the important result that there is a significant decrease in throughput capacity per node as node density increases. However, the analysis does not capture routing-related effects and assumes that all paths

Collision domains and nominal capacity

The model presented in [8] uses nominal MAC throughput and the concept of collision domains for estimating capacity of wireless mesh networks. The former is defined as the throughput achieved at the MAC layer in a one-hop IEEE 802.11 network operating in infrastructure mode. It depends on a number of factors such as MAC layer characteristics, channel conditions, network topology and packet size distribution etc. and can be determined if the relevant parameters are known [13].

Collision domains

Analytical model

The method described above for calculating the capacity of a chain network can be extended to arbitrary topologies [12]. Consider, once again, a network of N wireless nodes represented by the graph G(V, E) where V is the set of nodes and E is the set of links between these nodes. Each node is assumed to have p radio interfaces and q available channels. An arbitrary channel assignment algorithm allocates channels to radio interfaces of each network node. One of the network nodes is designated as

Scenarios for capacity analysis

Design of WMNs has to take into account a number of factors such as network architecture, size and topology, expected traffic profile, radio characteristics, channel assignment scheme, routing algorithm etc. Each of these factors has a bearing on the capacity of the resulting network. Therefore, it is vital that the capacity of the proposed network is estimated before it is deployed. Although capacity is just one of the parameters that may influence the performance of the WMN, it is

Capacity evaluation

The capacity of WMNs corresponding to the scenarios presented above was estimated using the analytical model presented in Section 4. For this analysis, the WMN nodes were assumed to have two radio interfaces. Furthermore, two values were used for the number of available channels—3 and 12. The former corresponds to the number of non-overlapping channels in IEEE802.11b while the latter equals the number of channels available in IEEE802.11a. A 5×5 grid is considered here. All the aggregator nodes

Conclusions

In this paper, we presented an analytical framework for determining the nominal capacity of multi-radio multi-channel wireless mesh networks. The objective of this work was to study the impact of various network design parameters on WMN capacity. The notion of collision domain is used to calculate an upper bound on the capacity available to ingress nodes that generate traffic towards the node that acts as the WMN gateway to the outside world.

The proposed analytical method was used to estimate

References (13)

  • I.F. Akyilidz et al.

    Wireless mesh networks: a survey

    Elsevier Comput. Netw. J.

    (2005)
  • P. Kyasanur, N. Vaidya, Routing and Interface assignment in multi-channel multi-interface wireless networks, in Proc....
  • P. Gupta et al.

    The capacity of wireless networks

    IEEE Trans. Inf. Theory

    (2000)
  • J. Li, C. Blake, D.S.J.D. Couto, H.I. Lee, R. Morris, Capacity of ad hoc wireless networks, in Proc. of 7th ACM...
  • M. Grossglauser et al.

    Mobility increases the capacity of ad hoc wireless networks

    IEEE/ACM Trans. Netw.

    (2002)
  • M. Gastpar, M. Vetterli, On the capacity of wireless networks: the relay case, in Proc. of IEEE INFOCOM 2002, vol. 3,...
There are more references available in the full text version of this article.

Cited by (0)

View full text