Restore

Network Characteristics

Our definition of the architecture of a WMN leads to several characteristics. These are quite general and many of them also hold for other perceptions of WMNs. To a certain extent they are also valid for (hybrid) MANETs [RufStuCla+05], but there are more or less subtle differences.

  • Wireless: The most obvious property is the wireless nature of WMNs. Thus, WMNs must cope with the challenges that arise from wireless communication. On the one hand, they need to take into account the limited transmission range and the potentially high loss rates due to packet collision and fading of the wireless channel during the transmission. On the other hand, they have to deal with the mobility of nodes.
  • Multi-hop: WMNs use multi-hop routing to overcome the challenges mentioned above. Conventional wireless networks extend their network coverage by higher transmission power or additional access points that have to be interconnected by wire. In contrast, nodes of a WMN forward traffic wirelessly on behalf of others, which are not within their direct transmission range.
  • Redundancy: The wireless backbone of a WMN forms a meshed network. It provides redundant links between mesh routers, mesh gateways, and mesh clients. Thus, failure of one link or node will not necessarily lead to failure of large parts of the network. Trying to adopt this approach in conventional wired networks might be expensive, time-consuming, or even impossible. This is because of the large amount of cabling required for such a meshed network. Depending on the environment, in-wall installations could be impossible.
  • Mobility: Since both mesh routers and mesh gateways have low mobility, the backbone can support client mobility in a predictable and reliable fashion. Mobile clients that leave the communication range of one mesh router can easily connect to the next one coming into their communication range. The dynamic multi-hop routing will ensure that the traffic is still correctly forwarded to its destination.
  • Dynamics: All nodes have to establish the network spontaneous way (self-organizing) and to maintain their connectivity continuously (self-healing). Leaving or newly joining nodes cause topology changes that the network has to adapt to. Nodes must reorganize their routes, invalidate paths that are not available anymore and include new paths that have become available. Additionally, the WMN should pass configuration information to new nodes in order to reduce or remove the need for user intervention (self-configuring).

The characteristics mentioned above are essential, but MANETs share similar qualities. The following characteristics are specific to WMNs, they clearly distinguish WMNs from MANETs, even from hybrid ones.

  • Infrastructure: Unlike MANETs, WMNs have got a pre-existing, hierarchical architecture. Mesh gateways and routers forming the backbone infrastructure are nearly static and therefore less limited with regard to power consumption and computing power. They can be equipped with multiple radios, and they can take over routing and configuration tasks. The hierarchical static infrastructure simplifies such functionalities since the backbone is more reliable than mobile nodes are.
  • Integration: The non-routing mesh clients can join a WMN without the need for sophisticated routing support. Thus, light-weight and power-constrained clients can be attached to a WMN. They need not be an active part, in contrast to MANETs that require all nodes to be cooperative. Supporting such passive clients enables the integration of devices, or whole networks, into the wireless backbone.

The latter characteristics influence the ones mentioned before. The introduced hierarchy adds several improvement opportunities in comparison to MANETs.