WO2009068586A2

WO2009068586A2 - Performance enhancing wireless network configuration

Info

Publication number: WO2009068586A2
Application number: PCT/EP2008/066276
Authority: WO
Inventors: Shahid Chaudary; Kevin Hicks
Original assignee: Tesco Stores Ltd
Priority date: 2007-11-26
Filing date: 2008-11-26
Publication date: 2009-06-04
Also published as: GB2455074A; WO2009068586A3; GB0723109D0

Abstract

This invention relates generally to data communications networks, and more specifically to methods and apparatus for enhancing the performance of networks which include wireless access points. The invention includes a data communications network comprising a first node of a first type configured as a wireless connection point, a second node of a first type configured as a wireless connection point, a node of a second type, different from the first type, a non- wireless connection between said first node of the first type and said node of the second type, a wireless connection interconnecting said first node of the first type and said second node of the first type, and means for directing traffic of a first type via said non-wireless connection and for directing traffic of a second type, different from said first type, via said wireless connection.

Description

Performance Enhancing Wireless Network Configuration

Field of the Invention

The present invention relates generally to data communications networks, and more specifically to a method and apparatus for enhancing the performance of networks including wireless access points.

Background of the Invention

Data communications networks often include elements that are connected by a wireless link. There are many benefits to a wireless connection, in particular providing the mobility of a wirelessly connected device. Typically, a number of static wireless connection points may be deployed within a local area to which mobile devices may form temporary wireless connections, the connections typically conforming to an industry standard protocol such as IEEE 802.11. Preferably the wireless connection points are arranged in such a way to give useful coverage in the area concerned and are connected to a data network by wired or wireless links. However, the wireless bandwidth available is typically subject to limitations due to spectrum allocation imposed by regulatory authorities and so it is desirable to ensure that the wireless resources are used in an efficient manner. This is typically achieved by careful planning of the siting, transmission power and frequency of operation of the wireless connection points and by the imposition of an efficient medium access control protocol.

A conventional arrangement of wireless connection points with overlapping wireless coverage needs to be planned and set up by an expert installer, which is time consuming and costly. Conventionally, interference between wireless connection points is avoided by the use of non-overlapping frequency channels (typically a choice of three channels for IEEE 802.1 Ib) with re-use allowed where path loss is sufficiently high to prevent co-channel interference. The transmission power and frequency are set up by the installer for each wireless connection point; if a wireless connection point fails or propagation conditions change (e.g. by the movement of objects into or out of the propagation path) then re-confϊguration also requires the skills of an expert installer.

It is possible to avoid the complications of planning frequency allocations by the use of a single frequency channel for the whole wireless network, with an appropriate medium access control protocol to prevent interference between links. A conventional implementation of a single frequency network involves setting up access points as wireless bridges so as to form a mesh.

Such a single frequency mesh network is described a paper by Bicket et al. entitled "Architecture and Evaluation of an Unplanned 802.11b Mesh Network" (MobiCom'05). A further example of a mesh network is given in patent application WO 2006/029126.

Figure 1 shows a conventional wireless mesh network in diagrammatic form. Only 3 nodes with wireless connections are shown here, 11a, l ib and l ie. It can be seen that some of these nodes (Ha and l ie in this case) have wired connections 13a, 13b in addition to wireless connections. Typically a minority of nodes have wired connections; in the mesh network described by Bicket et al, thirty-seven nodes share four wired connections. The wireless connections typically carry both data traffic 15 (i.e. payload traffic) and traffic related to the management of the wireless mesh 16 (i.e. non-payload traffic), typically multiplexed onto the same wireless connection. Node 12 has wired connections only and may be connected to other networks.

Figure 2 shows a wireless mesh network such as is described in WO 2006/029126. Nodes 20a and 20b are configured as wireless connection points, which are points in the network to which devices equipped with wireless transceivers of an appropriate type can make connections, typically of a temporary nature. Node 21 is configured as a wireless access point, i.e. a type of wireless connection point with a wired connection. In this case, the wired connection 23 is to a non-wireless node 24 which may be connected to other nodes. Both data 26 and management 25 traffic (i.e. payload and non-payload traffic respectively) are carried on the wireless connections between nodes 20a, 20b and 21. Mobile data units 22a, 22b and 22c have made wireless connections to the wireless connection points. In this example, nodes 20a and 20b are configured as non-root wireless bridges and node 21 is configured as a root wireless bridge. A wireless bridge is a type of wireless connection point that does not connect traffic to a wired connection. The significance of the designation as a root wireless bridge is that the bridge controls the operation of the wireless mesh network, setting, for example, the operating frequency and determining the topology of the connections between nodes. The non-root wireless bridges act as repeaters, so that data is passed from one to the next over the wireless network. The data flow in the network can be illustrated by the following example: data from mobile data unit 22c is passed via node 20b to 20a and on to the root wireless bridge 21 before connection to the wired network. The limitation of this approach is that data capacity is used by the repeating function. Conventional methods of control of access to radio resources will now be considered, for isolated wireless connection points, mesh networks and wireless connection points with overlapping coverage.

Taking as an example the IEEE 802.11b wireless standard, the default medium access control (MAC) protocol that is conventionally used is carrier sense multiple access with collision avoidance (CSMA-CA). This is quite serviceable and robust for areas covered by isolated wireless connection points or ad hoc networks, but it is not optimum for efficiency for mesh networks or for networks of wireless connection points with overlapping coverage as will be explained. The CSMA-CA protocol operates by testing for the presence of carriers from other signals before transmission, and only transmitting if the channel is clear. If a carrier is sensed, then the terminal waits for a randomly determined period before retransmitting. The random time delay prevents repeated clashes; one of the parties clashing should be delayed relative to the other on a second attempt and so its carrier should be detected and transmission from the other party held up, preventing a second clash. However, the CSMA- CA used alone is not as efficient a medium control protocol as can be achieved by a coordinated approach between adjacent links. In addition, the CSMA-CA protocol can suffer from the "hidden node" problem, by which, due to signal attenuation, a transmitter does not see an interfering carrier which is nevertheless strong enough to cause interference at the intended recipient. More efficient use of the radio resource can be provided by coordination between potentially interfering wireless links, to prevent clashes occurring and removing the need for a randomly determined back-off delay before retransmission. A protocol of this type is the point coordination function (PCF) which can be used for controlling the access to the radio resource within the coverage area of a master unit, typically a wireless connection point. However, this protocol as defined in IEEE 802.11 is not efficient at coordinating radio resource control between the coverage of overlapping wireless coverage areas, such as that of access points as pointed out by Vergados et al in a paper entitled "Synchronisation of multiple access points in the IEEE 802.11 point coordination function" (IEEE 60^th VoI 2, 26-29 Sep 2004).

The approach taken to medium access control in a wireless mesh depends upon the standard to which it conforms; wireless meshes may typically be configured to a proprietary standard requiring that nodes within the network conform to that standard. In some cases CSMA-CA may be used alone and in other cases the CSMA-CA protocol may be supplemented by coordination of access between nodes, for example by the conveyance between nodes of ready- to-send/clear-to-send (RTS/CTS) signals between nodes. Such signals will typically be conveyed between nodes in the form of management traffic multiplexed with other payload traffic using the wireless connection between nodes. However, as pointed out above, the efficiency of a wireless mesh network in conveying data is limited by the need to relay data between nodes.

Rather than setting up a single frequency mesh, it is possible to configure a single frequency network, in which each wireless connection point has a wired connection to further portions of a network as illustrated in Figure 2a. In this case, in contrast to a wireless mesh network, there is no direct wireless link between wireless connection points 10a, 10b, 10c. As has been mentioned, wireless connection points with a wired connection to further portions of a data network can be referred to as access points. In order to coordinate the MAC between access points with overlapping coverage to provide a medium access control means in addition to CSMA-CA, it is necessary to include a form of communication between the access points. Typically this is carried out via the wired connections 13a, 13b, 13c, 14 from access points 10a, 10b, 10c to a centralised controller, such as a proprietary wireless management controller 5, which may be expensive and complex and is typically shared between several local area networks and sited at a remote location The link 14 to the central office 8 may be a leased line with lower data capacity than other links in the network. It should be noted that non-payload traffic related to the management of access to the wireless resource 6 is present on the link 14 to the central office in addition to payload traffic.

It is an object of the present invention to provide a method and apparatus which addresses these disadvantages.

Summary of the Invention

In accordance with a first aspect of the present invention, there is provided methods and apparatus according to the appended claims. The nodes of the first and second type may for example include routeing or bridging functions and will typically include transceivers appropriate for a wireless or non- wireless connection to another node or nodes. Preferably, the traffic of the first type is data traffic that passes via at least one node of a first type and one node of a second type and preferably traffic of the second type is management traffic that is confined to a network including only the wireless connections between nodes of the first type. Traffic of the first type can be referred to as payload traffic (for example voice data or video data), whereas traffic of the second type can be referred to as non-payload traffic.

Wireless connections may conform to an industry standard such as IEEE 802.11 or IEEE 802.15 or it may be a proprietary or other format or may operate in a variety of alternative formats. Wireless links include free-space optical links. Non- wireless connections include, but are not limited to, links utilising a wired, power line carrier or optical fibre medium conforming for example to an Ethernet or other standard or a proprietary or other format or a variety of alternative formats. The means for directing traffic may be a device or a combination of devices and may have functions other than solely directing traffic. For example, the means for directing traffic may be a configuration of a proprietary wireless connection point or access point. A benefit associated with the apparatus of claim 1 is that traffic capacity of the non- wireless connection is not taken up by management traffic and is accordingly available for use by data traffic.

Preferably, data traffic (i.e. payload traffic) includes messages received by or transmitted from a node of the first type via a wireless connection to and from a node not configured as a wireless connection point. For example, the wireless connection may be a temporary or permanent association formed between a mobile or static wireless device and a wireless connection point, with the benefit of control of traffic such that the traffic on the connection is directed via a wired link rather than a wireless connection. For example, the payload traffic may relate to stock taking in a retail establishment or warehouse.

Conveniently, management traffic (i.e. non-payload traffic) includes messages related to the control of access to the wireless resource. The access to the wireless resource may be by nodes of the first type or by other nodes with a wireless connection such as for example mobile wireless devices. For example, the traffic may relate to the set up and maintenance of wireless connections or the coordination of medium access control between nodes to ensure efficient use of the wireless resource.

A particular type of management traffic is mesh management traffic. Mesh management traffic includes messages related to establishing and maintaining the mesh network, and may relate to channel frequency, received signal strength, interference level, signal to noise ratio, MAC address and other information relating to neighbouring nodes, association tables (that is, which mobile data units are associated with a node), and other parameters related to the operation of the mesh network.

Advantageously, the nodes of a first type are configured as a wireless mesh network. A wireless mesh network may conform to an industry standard such as the IEEE 802.11 or IEEE 802.15 standards mentioned above or may be a feature of a proprietary system or may be an arrangement configured according to a bespoke standard. In principle, a wireless mesh network operates such that each node may communicate with any other node within the mesh by means of communication across one or more wireless links. The benefit of configuring the nodes of a first type as a wireless mesh is that such a configuration may provide coordination of access to the wireless medium providing efficient medium access control.

Preferably, a data communications network as described above is configured as one or more virtual area networks, each of which consists of a defined sub-set of network elements. Conveniently, a first virtual local area network is configured to include the nodes of the first type and also the node of a second type. A second virtual local area network is configured to include only the nodes of the first type connected only by wireless connections. Payload traffic is associated with the first virtual local area network and non-payload management traffic is associated with the second virtual local area network. These associations enable a convenient means of directing traffic, so that management traffic does not occupy capacity on wired connections. So, for example, the first VLAN may include wireless connection points, mobile wireless devices and other nodes and the connections directly between them, whereas the second VLAN may be confined to comprising the wireless connection points and the connections directly between them.

Conveniently, a data communications network is provided as described above wherein wireless connections within the first virtual local area network are allocated a high path cost and non-wireless connections within the first virtual local area network are allocated a low path cost such that payload traffic is directed via non- wireless connections in preference to wireless connections. Path cost is a parameter that can be associated with a link between nodes in a network. Parallel paths between nodes are undesirable in some cases, so to avoid the creation of parallel paths a procedure is implemented that selects paths with the lowest total path cost for a route between given nodes in a network. An example of such a procedure is the spanning tree protocol. So, the allocation of a high path cost to a link when an alternative link or combination of links is available with a lower associated path cost can be used to select a preferred route via a physical connection.

Advantageously, the wireless connections operate in a single frequency channel, with the benefit that planning and installation are potentially simplified. Taking the IEEE 802.11b standard for example, where thirteen potential frequency channels are defined for use in Europe and in a single frequency network each node would be configured to operate on the same one of these frequency channels. It may be the case that nodes are capable of operating within other frequency channels in addition to the single frequency selected for communication within the network.

A further aspect of the invention relates to a node in a data communications network, the node being equipped with a wireless connection to and from a second node and a non-wireless connection to and from a third node. The node is equipped with a traffic director which ensures that payload traffic is conveyed by the wireless connection and management traffic is conveyed by the non-wireless connection. The benefit given by the use of a node of this type is that of control of the loading of traffic onto connections utilising different media. Such connections may differ in terms of capacity to carry data and so the control of traffic loading may be valuable in optimising use of the network.

A further aspect of the invention relates to a method of controlling the flow of traffic to and from a first node of a data communications network in which the node is equipped with a wireless connection to and from a second node and a non-wireless connection to and from a third node. This is achieved by associating the first node with two or more virtual local area networks. However, the port on the first node that is associated with the non-wireless connection is associated with a first virtual local area network and not with a second virtual area network. As a result, the port will not convey data associated with the second virtual local area network, provided that the port is configured as a non-trunked port, that is to say that a trunking function is not enabled on that port. A trunking function is a method by which traffic from more than one virtual local area network can be multiplexed onto a connection between nodes involving tagging of data frames with an identifier relating to the appropriate virtual local area network. As a result of not enabling the trunking function, traffic associated with the second virtual local area network is prevented from being conveyed by the non-wireless connection associated with the first port. The benefit of this method of controlling the flow of traffic is that it may be implemented using a configuration of an existing network node without the requirement for further hardware.

Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the invention, given by way of example only, which is made with reference to the accompanying drawings.

Brief Description of the Drawings Figure 1 is a diagram showing a conventional wireless mesh network;

Figure 2 is a diagram showing a conventional wireless mesh network comprising wireless access points and wireless connection points;

Figure 2a is a diagram showing a conventional single frequency network comprising wireless access points; Figure 3 is a diagram showing a first embodiment of the invention;

Figure 4 is a diagram showing a second embodiment of the invention; Figure 5 a is a diagram showing part of the embodiment of Figure 4 illustrating a first virtual local area network;

Figure 5b is a diagram showing part of the embodiment of Figure 4 illustrating a second virtual local area network; and Figure 5 c is a diagram showing part of the embodiment of Figure 4 illustrating the traffic direction means.

Detailed Description of the Invention In general, the present invention is directed to methods and apparatus that enhance the performance of data networks that include wireless connections. For clarity, the methods and apparatus are described in the context of a wireless data capture system, such as are used in retail establishments and warehouses. However, it is to be understood that the invention is not limited to such an application. For example, the present invention may be applied to data networks used to communicate any type of data including but not limited to digitally encoded voice signals, audio signals generally, images and video streams. The wireless signals may conform to industry standards such as IEEE 802.11 but could also conform to other industry standards such as ultra- wideband radio, to proprietary standards or could conform to no particular standard. The wireless signals may be radio frequency electromagnetic radiation, free-space optical signals, or signals transmitted or received through any medium or free space in a wireless manner.

Embodiments of the invention relate to ways of managing traffic in a network including wireless access points to maximise payload capacity without the use of a controller at a central office. Conventionally, management of access to wireless resources is controlled from a central office that may be some distance from a network of wireless access points; this is so that several wireless networks at different locations may share a single controller in order to save costs. However, the capacity of the links back to the central office from a wireless network may be a limiting factor in the payload capacity of the network; the links may be leased lines in which the bandwidth has been minimised in order to save costs.

According to embodiments of the invention, non-payload traffic such as traffic relating to management of access to the wireless resources is preferentially routed via wireless connections between access points, rather than using the wired links between the access points and a central office, thus releasing capacity on wired connections for payload traffic. This removal of non-payload traffic from the wired links between the access points and the central office may be conventionally achieved by implementing a meshed network between the wireless access points; management of access to the wireless resource is self-contained within the meshed network. However, a conventional meshed network is implemented by routing payload traffic over wireless links between access points; this is undesirable if each access point has a wired link, since it is wasteful of wireless resources, thus reducing the capacity of the wireless links from user equipment to the access points, and also can add latency by the addition of unnecessary wireless hops.

Therefore, according to embodiments of the invention, payload data received by user equipment connected to an access point is preferentially routed via the wired links from the access point and is excluded from the wireless links between access points. Hence, a meshed network may be set up between access points, which has the beneficial effect of managing under local control the access to the wireless resource for the links between access points and user equipment. Payload traffic is excluded from the mesh, so that wireless resources may be reserved for links between access points and user equipment. This enables efficient management of access to the radio resources whilst saving the cost of a controller at the central office, and in cases in which there is a bottleneck in the backhaul link from the wireless access points to a central office, results in an improvement in payload capacity of the backhaul network.

By way of example a specific embodiment of the invention will now be described in detail in the context of an area such as business premises, which is provided with a network of wireless connection points to which one or more mobile data units can form wireless connections. The connection points may for example include routeing or bridging functions and will typically include transceivers appropriate for a wireless connection and also a wired connection to a further portion of a data network, which may be a corporate network including a data centre in another location or it may include a connection to the internet. The purpose of the provision of connection points may be for stock taking or other purposes, requiring that the mobile data unit, which may be a PDA (personal digital assistant) can be moved within the wireless coverage area of a connection point and also between connection points within the business premises. Other types of device equipped with a wireless transceiver can of course also be connected to the network via the connection points, such as personal computers (PCs).

For reasons of network security, there is typically an authentication process involved before data can be transferred between the mobile data unit and the wireless connection point. One method of authentication involves the provision of a code word by the mobile data unit such as a Wired Equivalent Protection (WEP) key. It may also be necessary for the mobile data unit to request connection to the correct Local Area Network (LAN). Connection points provided by some vendors provide for roaming, that is to say a mobile data unit may move between areas covered by different connection points with the association to one connection point being shut down and the association to the next connection point being set up by an automated process. However, it may be necessary for the mobile data unit to provide authentication code words before the connection to a different connection point is established. Typically, the choice of connection point to which an association is set up is made on the basis of received signal strength.

An embodiment of the invention in a network including three wireless connection points is shown in Figure 3. Wireless connection points 30a, 30b and 31 are connected by a wired link to a non-wireless node 37, which may itself be connected to a wider data network. A mobile data unit 32a is connected by a wireless link to the first wireless connection point 31 , a second mobile data unit 32b is connected by a wireless link to a second wireless connection point 30b and a third mobile data unit 32c is connected by a wireless link to a third wireless connection point 30b. The medium access control on the wireless links between mobile data units and wireless connection points is coordinated by means of management traffic on wireless links 35 and 36 between the wireless connection points. A traffic director is provided such that management traffic relating to wireless links is preferentially carried on the wireless links 35 and 36 whereas other traffic such as payload traffic between the mobile data unit 32a or 32b or 32c and node 37 is preferentially carried on a wired link. The traffic director may be a device or a combination of devices and may have functions other than solely directing traffic. For example, the traffic director may be configured as a proprietary wireless connection point. The traffic director may distinguish between traffic of different types on the basis of information contained in a designated part of the data structure of a frame, such as the data header. As a result of the operation of the traffic director, payload traffic between the mobile data units 32a, 32b and 32c and node 37 is prevented from occupying bandwidth on wireless links 35 or 36. Also, traffic related to management of access to the wireless resource (non-payload data) is prevented from following a path including wired links. Figure 4 shows an embodiment of the invention in a system that comprises a larger number of nodes than are shown in Figure 3. In this example, six nodes 40a, 40b, 40c, 4Od , 4Oe and 41 are configured in such a way that they are able to act as access points and are referred to here as hybrid nodes. The hybrid nodes are connected by wired links to an Ethernet switch 43, which is connected to a back office management and authentication functional block 44. Six mobile data units 42a, 42b, 42c, 42d, 42e and 42f, in this example mobile PDAs, are shown connected to the hybrid nodes. In the case of hybrid node 4Od, two mobile PDA units are connected by wireless links, illustrating that there may be more than one mobile PDA connected to each hybrid node. The configuration shown in the example of Figure 4 illustrates the case where hybrid nodes 40a, 40b, 40c, and 4Od are connected directly to node 41, but it is not a requirement that all hybrid nodes are connected to node 41, and indeed hybrid node 4Oe is not connected directly by wireless link to hybrid node 41; rather, it is connected via node 4Od. Wireless connections 45a, 45b, 45c, 45d and 45 e are configured as a wireless mesh network. All wireless connections in this embodiment operate at the same frequency channel, designated arbitrarily as fi. Each hybrid node includes a traffic director functional block, the implementation of which is related to the operation of virtual local area networks; the network of Figure 4 is partitioned into two virtual local area networks as illustrated in Figures 5 a and 5b. A virtual local area network (VLAN) is a layer 2 entity; that is, it functions at the medium access control (MAC) layer of the industry standard OSI seven layer telecommunications model. Devices within a VLAN can communicate with other devices within the same VLAN without communicating through a router. One conventional function of a VLAN is to implement logical segregation, so that devices that are members of one VLAN cannot communicate with devices that are members of other VLANs (unless enabled to do so by a router). An advantage of this is that the amount of broadcast traffic is limited, as broadcast traffic on one VLAN is not seen by members of another VLAN. Certain ports of a device such as an Ethernet switch or access point can be made members of a VLAN. So it can be seen that if traffic is carried within a given VLAN, it can be directed to certain connections between nodes that are associated with ports that are members of the given VLAN and not to connections whose ports are not members of that VLAN.

Conventionally, it is possible to institute a trunking function such that traffic belonging to two or more VLANs can be carried by a single port and a single connection. This is achieved by tagging data frames to indicate the VLAN membership of a frame so that the data from two of more VLANs can be multiplexed together and separated upon reception. Suitable trunking protocols include IEEE802.11q and the Cisco proprietary Inter-switch linking (ISL) protocol. However, if trunking is not implemented, then the allocation of a port on a device to a VLAN effectively blocks data frames that are carried on a different VLAN from using that port, which is referred to as a non-trunked port. The non-use of trunking is exploited for use in traffic direction in preferred embodiments. It is possible for a network element to be a member of two or more

VLANs in such a way that the node is configured differently as regards the operation in association with each VLAN. For example, in this embodiment a wireless connection point is configured as what has been referred to as a hybrid node. The description of the node as a "hybrid" relates to its acting as a wireless access point when operating as a member of one VLAN and as a wireless bridge when operating as a member of another. As stated above, the distinction is that a wireless access point has a wired connection to another node, whereas a wireless bridge is set up to route data from wireless node to wireless node in one or more hops to a root bridge, at which point the data is connected to a wired connection. "Wireless connection point" is used as a generic term to include both "wireless access point" and "wireless bridge".

In the network of Figure 4, as has been stated, two VLANs are set up. Figure 5a shows the network elements that are members of a first VLAN, VLANl. It can be seen that VLANl includes mobile PDAs 42a, 42b, 42c, 42d, 42e and 42f, hybrid nodes 40a, 40b, 40c, 4Od, 4Oe and 41 configured as access points for VLANl, Ethernet switch 43, and the ports connected to the wired connections between the access points and the Ethernet switch. In addition, nodes in a further network portion beyond that illustrated in Figure 5a may be members of VLANl. Note that payload traffic to and from a mobile PDA will be preferentially allocated to VLANl and so is not carried by the other VLAN. Figure 5b shows the network elements that are members of a second

VLAN, VLAN20. It can be seen that VLAN20 comprises the hybrid nodes 40a, 40b, 40c, 4Od, 4Oe and 41, connected as a wireless mesh, and the wireless connections between them. A root node, 41, controls the frequency channel in which the mesh network operates. The setting up of a mesh network may be a pre-programmed function of a proprietary access point and may involve, as has been mentioned, the setting of the operating frequency of the mesh, and the establishment of links between the nodes. Not all possible links between nodes will be established; if alternative routes are available then one will be chosen, typically on the basis of wireless signal quality. Consequently, information relating to wireless signal quality, for example received signal strength, forms part of the management traffic on the wireless mesh network. To establish efficient medium access control to the wireless resource, it is necessary to coordinate medium access between nodes that are members of the mesh since the coverage of the nodes will in many cases overlap. This is carried out within the mesh network by signalling across the wireless links. Ready To Send/Clear To Send (RTS/CTS) signals are an example of a method of coordinating medium access control between wireless links; other types of signalling could be used.

It will therefore be appreciated that the wireless mesh is not intended to carry payload traffic itself; it is set up for the purpose of coordinating medium access control in links that do carry data and that connect nodes that are outside the mesh to nodes that are part of the mesh, for example mobile data units. In this respect, VLAN20, configured as the mesh network, has the unusual characteristic that it typically carries no data except for management traffic.

It is to be noted that typically, mobile data units such as PDAs will not be part of the wireless mesh, but can form associations with it and communicate with the nodes that are part of the wireless mesh. In forming the associations with the wireless mesh network, the wireless links between the mobile data units and the nodes which are part of the wireless mesh network are also subject to the coordinated medium access control that is provided by the mesh network; this is a benefit of setting up the wireless mesh.

Figure 5 c shows how traffic is directed within a hybrid node in one embodiment. Considering firstly traffic associated with VLANl, payload traffic between mobile PDA 42a and hybrid node 41 is preferentially carried by a wireless link. The wired link 51a to and from the hybrid node 41 is allocated a low path cost, whereas potential wireless path 46 is allocated a high path cost. A spanning tree protocol or other protocol is operated such that the low path cost route 51a is selected rather than the high path cost route 46. The payload traffic is accordingly preferentially directed to the wired link rather than the wireless link. Considering next traffic associated with VLAN20, as the wired port 50a on hybrid node 41 is allocated to VLANl and trunking is not implemented as discussed above, this port is blocked to traffic on VLAN20. Traffic associated with VLAN20 is thus preferentially confined to the wireless connection rather than taking a route via wired connections. Any traffic on VLAN20 destined for nodes which would require connection via wired connection 51a to a node outside the wireless mesh is therefore blocked and can be said to be routed to null, as shown in connection 49. In practice, as has been stated, it is not intended for VLAN20 to carry traffic except for non-payload management traffic. Therefore the cases where traffic associated with VLAN20 is destined for a node outside the wireless mesh are rarely occurring and may be due to a mobile PDA erroneously connecting to VLAN20 due to a configuration error or due to unauthorised use. Such traffic would be routed across the wireless mesh to root node 41, where, as shown, it would in any event be routed to null, that is to say, the data is lost.

It has thus been shown that the embodiment of Figure 4 makes use of a wireless mesh network in relation to the coordination of access to the wireless medium and also makes use of the configuration of virtual local area networks in combination with the allocation of path cost functions to direct the flow of data on the network.

The above embodiments are to be understood as illustrative examples of the invention. It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the embodiments. Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the invention, which is defined in the accompanying claims.

Claims

1. A data communications network comprising: a first node of a first type configured as a wireless connection point; a second node of a first type configured as a wireless connection point; a node of a second type, different from the first type; a non-wireless connection between said first node of the first type and said node of the second type; and a wireless connection between said first node of the first type and said second node of the first type; means for preferentially directing payload traffic via said non-wireless connection, said payload traffic being transceived by said first node of the first type via a wireless connection to or from a further node, said further node not being configured as a wireless connection point; and means for preferentially directing non-payload traffic via said wireless connection.

2. A data communications network according to claim 1, wherein said further node comprises a mobile data unit.

3. A data communications network according to claim 1 or 2, wherein said non-payload traffic comprises control data relating to control of access to wireless resources used by said first node of the first type configured as a wireless connection point.

4. A data communications network according to any preceding claim, wherein said payload traffic does not comprise control data relating to control of access to wireless resources used by said first node of the first type configured as a wireless connection point.

5. A data communications network according to any preceding claim, wherein said non-payload traffic comprises control traffic for use in configuring a wireless mesh network comprising nodes of the first type.

6. A data communications network according to any preceding claim, wherein said first and second nodes of said first type are connected with other nodes of the first type as a wireless mesh network.

7. A data communications network according to any preceding claim, comprising a first virtual local area network and a second virtual local area network, said first virtual local area network being configured to include said first and second nodes of the first type and said node of a second type and said second virtual local area network being configured to include said first and second nodes of the first type interconnected by wireless connections, wherein payload traffic is associated with said first virtual local area network and non-payload traffic is associated with said second virtual local area network.

8. A data communications network according to claim 7, wherein said wireless connections within said first virtual local area network are allocated a high path cost and said non-wireless connections within said first virtual local area network are allocated a low path cost such that payload traffic is directed via a non-wireless connection in preference to the wireless connections.

9. A data communications network according to claim 7 or 8, wherein the wireless connections operate in a single frequency channel.

10. A node of a first type in a data communications network, the node being configured as a wireless connection point and the node comprising: a first port providing a wireless connection to and from a second node; a second port providing a non-wireless connection to and from a third node; and a traffic director operable to direct traffic such that traffic of a first type is conveyed by said wireless connection and traffic of a second type, different from the first type, is conveyed by said non-wireless connection.

11. A node according to claim 10, wherein the traffic director is arranged to control transmission of traffic of the first type via a wireless connection to and from a node not configured as a wireless connection point.

12. A node according to claim 10 or claim 11, wherein the traffic director is arranged to control traffic related to control of access to the wireless resource via said non- wireless connection.

13. A node according to any one of claim 10 to claim 12, wherein: the node is associated with both a first virtual local area network and a second virtual local area network, traffic of said first type is associated with said first virtual local area network, traffic of said second type is associated with said second virtual local area network, the second port is associated with the first virtual local area network, and the second port is configured as a non-trunked port such that said traffic of said first type may pass via said non-wireless connection and said traffic of said second type is prevented from passing via said non-wireless connection.

14. A node according to claim 13, wherein said wireless connection via said first port within said first virtual local area network is allocated a high path cost and said non-wireless connection via said second port within said first virtual local area network is allocated a low path cost, such that traffic of said first type is directed via the said non-wireless connection in preference to said wireless connection.

15. A method of configuring a data communications network comprising a first node of a first type, a second node of a first type and a node of a second type, different from the first type, the method comprising: configuring the first node a wireless connection point; configuring the second node as a wireless connection point; connecting the first node to and from the node of the second type via a non- wireless connection; connecting the first node to and from the second node via a wireless connection; preferentially directing payload traffic via said non-wireless connection, said payload traffic being transceived by a node of the first type via a wireless connection to or from a further node, said further node not being configured as a wireless connection point; and preferentially directing non-payload traffic via said wireless connection.

16. A method according to claim 15, wherein said further node comprises a mobile data unit.

17. A method according to claim 15 or 16, wherein said non-payload traffic comprises control data relating to control of access to wireless resources used by said first node of the first type configured as a wireless connection point.

18. A method according to any of claims 15 to 17, wherein said non- payload traffic comprises management traffic for use in configuring a wireless mesh network comprising nodes of the first type.

19. A method according to any of claims 15 to 18, wherein said payload traffic does not comprise control data relating to control of access to wireless resources used by said first node of the first type configured as a wireless connection point.

20. A method according to any of claims 15 to 19, wherein said first and second nodes of the first type are connected with other nodes of the first type as a wireless mesh network.

21. A method according to any of claims 15 to 20, further comprising: configuring a first virtual local area network to include said first and second nodes of the first type and one said node of the second type; configuring a second virtual local area network to include said first and second nodes of the first type connected only by wireless connections; associating payload traffic with said first virtual local area network; and associating non-payload traffic with said second virtual local area network.

22. A method according to claim 21, further comprising: allocating a high path cost to wireless connections within said first virtual local area network; and allocating a low path cost to non-wireless connections within said first virtual local area network such that payload traffic is directed via a non-wireless connection in preference to a wireless connection.

23. A method according to any one of claim 20 to claim 22, wherein the wireless connections operate in a single frequency channel.

24. A method of controlling the flow of traffic to and from a first node of a data communications network wherein the first node is configured as a wireless connection point and is equipped with a wireless connection to and from a second node and a non-wireless connection to and from a third node, and the first node is associated with both a first virtual local area network and a second virtual local area network, the method comprising: receiving data at the first node from a mobile data unit; associating said received data with said first virtual local area network; configuring a port of the first node with a non- wireless connection to the first virtual local area network; and configuring said port as a non-trunked port such that said first data is transmitted to said third node via said non- wireless connection.

25. A method according to claim 24, further comprising receiving second data at the first node from the second node and associating said second data with said second virtual local area network such that said second data is not transmitted via said non- wireless connection.