Method for control of communications from an edge device of an access network and edge device and network management module for performing said method

The present invention relates to method for communications control from an edge device of an access network, via the provisioning or the non-provisioning of at least one layer 2 destination address of at least one other edge node of this access network, to this edge device as is further described in the preamble of claim 1.

Such a method is already known in the art, e.g. by using pre-configured filters in access multiplexers of Ethernet access networks, wherein the allowed MAC-addresses of outgoing edge nodes are stored in these filters. The prior art method and system either use a pre-configured filter in the edge nodes themselves, or use a more centralised push-mechanism from where a central control server in the network management module provides each of the edge-nodes with their pre-configured lists of allowed network devices with which they may communicate.

In some access networks, for instance connectionless aggregation networks such as Ethernet access networks, failures or reconfigurations in this network after such failures are not known to the edge nodes. This means that, if a network failure takes place, and another MAC address is associated with the same layer 3 address of the other edge node, this information is not available to the filters in the ingress edge node. The existing push mechanisms which centrally keep track of these changes but only provide this information to the edge nodes from time to time are not dynamic enough to quickly signal the changes to the edge nodes such as the access multiplexers. The other mechanism whereby the filters in the edge nodes are preconfigured at start up does not provide a solution at all since with this method the changes are never known during the operation of edge devices such as the access multiplexers.

An object of the present invention is thus to provide a method for communications control from an edge device of an access network of the above known kind, but which is dynamic enough to adapt to changes in the layer 2 destination addresses due to unforeseen circumstances such as for instance network failures.

According to the invention, this object is achieved by the fact that these layer 2 destination addresses are only delivered upon request of the edge device itself, as is further stated in the characteristic part of claim 1 .

In this way, since the centralised network management module keeps track of the changes with respect to the allowed layer 2 destination addresses of edge devices within the access network, each time a request from an edge device is received with respect to such a communication to another edge device, the network management module performs an updated check and may send the latest known information to the edge node which accordingly has the up-to-date information available for its further communications.

Furthermore, the request is generated in the edge node upon arrival of a packet including the layer 2 destination address of the destination edge node in the access network, and upon checking whether this layer 2 destination address is not yet stored within a communications restriction filter in the edge node. This ensures that, for not yet locally stored destinations, always the latest information is obtained from the network management module.

It is to be noted that the US Patent US-A-5 968 176 by SHERER William Paul et al (19 October 1999) discloses a multilayer firewall system that provides for establishing security in a network that include nodes having security functions operating in multiple protocol layers. Therein is disclosed a method for control of communications from an edge device or node of an access network (column 3, lines 20-40), via provisioning or the non-provisioning of at least one layer 2 destination address of at least one other edge device of the access network to the edge device by a network management module of the access network (column 4, lines 21-26; column 5, line 4) as in the present invention. However, this known document doesn't disclose nor teach that, upon arrival of a packet including at least one layer 2 destination address, the edge device checks whether this layer 2 destination address is already stored within a communications restriction filter of the edge device and generates a request including this layer 2 destination address only in case this one layer 2 destination address is not yet stored within the communication restriction filter.

It is further to be noted that the International Patent Application WO-00 52575-A of "PACKET TECHNOLOGIES LTD" by DANIELY GAD (8 September 2000) also discloses a system for local, distributed security for a computer (22) connected to a network (42). The network (42) transmits packets to and from the computer (22). The system includes a local security device (20) for connecting the computer (22) to the network (42) and for examining each packet to determine whether the packet is received by the computer (22) according to one or more rules. Here again, even if the method disclosed in this other known document may be compared to the method for control of communications from an edge device or node of an access network as mentioned in the preamble of the present claim 1, this other known document also doesn't disclose nor teach that, upon arrival of a packet including at least one layer 2 destination address, the edge device checks whether this layer 2 destination address is already stored within a communications restriction filter of the edge device and generates a request including this layer 2 destination address only in case this one layer 2 destination address is not yet stored within the communication restriction filter. Moreover, this other known document doesn't mention anything about destination allowance as described above. Yet another characteristic feature of the present invention is described in claim 2.

This is extremely interesting in case of network failures where not only it will be known whether a desired MAC address can still be used or not for a destination, but by providing additional layer 2 forwarding information such as a VLAN tag, possibilities for service mapping and class of service segregation are provided. Furthermore by providing higher OSI layer information such as OSI layer 3 or OSI layer 4 information possibilities for protocol and application control can be provided. For example protocols and applications that are using a certain TPC port can be allowed or blocked. With these OSI layers reference is made to the well-known 7-layer OSI model in data communications of which layer 1 represents the physical layer, layer 2 the data layer, layer 3 the network layer, layer 4 the transport layer, layer 5 the session layer, layer 6 the presentation layer and layer 7 the application layer.

Still a further characteristic feature of the present invention is described in claim 5.

Thereby, ageing is introduced to the filters within the edge nodes which keep the allowed MAC addresses. This ensures again that on a regular basis the latest information is stored within the filters ensuring a dynamic communications control.

Yet a further characteristic feature of the present invention is described in claim 6.

By having the request containing additional user information with respect to the sender of the packet, possibilities for charging are opened, as is also stated in claim 7. This further allows the control of user-to-user communications within the access network itself, which are in general not allowed under normal operating conditions, since these are normally not stored within the network management module. By yet providing the possibility to deliver such MAC addresses upon consulting the charging device, as stated in claim 8, such user-to-user communications within the access network become now possible.

Claim 9 further states that the further passage of an incoming packet through the access network is blocked in case the layer 2 destination address, for instance the MAC address of that packet, is not stored within the edge device or not received by said edge device from the network management module.

The present invention also relates to an edge node and a communications restrictions module of a network management module which are able to perform the above described method, as respectively claimed in claims 10-15 and 16-21.

The above mentioned and other objects and features of the invention will become more apparent and the invention itself will be best understood by referring to the following description of an embodiment taken in conjunction with the accompanying drawings in which <ul id="ul0001" list-style="none" compact="compact"><li>Fig. 1 gives an overview picture of an access network AN with several edge devices, internal switches , a network management module, and other neighbouring networks in which a communication is desired from edge device ED 1 to edge device ED 2, and</li><li>Fig. 2 shows details of edge device ED 1 and the network management module NMM for performing the method according to the invention.</li></ul>

The present invention relates to a method for controlling and restricting communications to allowed edge devices in an access network. Such an access network, preferably a connectionless aggregation network such as an Ethernet access network, is depicted in Fig. 1. This figures shows 4 edge devices ED1 to ED4 of this network AN, of which ED1 and ED2 are access multiplexers for aggregation of traffic from several users which are depicted by the small houses coupled to these access multiplexers. Other edge devices such as ED3 and ED4 comprise interface devices with one or more service provider networks e.g. ED3 provides an interface to network service providers NSP B and NSP C, and ED4 provides an interface to network service provider NSP A.

Access networks are in general managed by a network management module which is depicted on Fig. 1 by block NMM. In general the task of such a network management module is to configure the access network by allocating bandwidths, charging, and user management possibilities. For the present invention one particular aspect of the network management is important, being its memory M1 for storing a list of allowed layer 2 addresses for corresponding allowed destination edge devices, for each edge device to which packets are entering the access network.

Looking more into detail to this network management module on Fig. 2, one can observe that, apart from this memory M1 where these lists for all allowed edge destination nodes and their corresponding layer 2 destination addresses are stored, the network management module includes a communications restriction module (CRM) which is able to retrieve from the memory M1 these layer 2 destination addresses, and delivers them to an edge node , for instance ED1, but only after having received a request from ED1 with respect to this particular destination address. To this purpose edge device ED1 includes a communications control device CDD which generates these requests. In some embodiments these requests are always generated upon arrival of a packet. In a preferred embodiment of the invention these requests are only generated upon arrival of a packet including this layer 2 destination address , and only after checking within said edge device (ED1) that this layer 2 destination address is not stored yet within a communications restriction filter (CRF) therein. This check is indicated in Fig. 2 by a first communication arrow from the CCD to the CFR with the message "MAC2?" , indicating that CCD informs whether MAC2 is present within CFR. The answer of CFR to CDD is indicated by the arrow from CRF to CCD with the message "noMAC2", indicating that this layer 2 destination address is not yet stored within the communications restriction filter. It may be remarked that, upon initialisation of the edge device, the filter may be preconfigured with some predetermined allowed addresses, or be empty.

If the communications control device CDD has thus not found this layer 2 destination address within this filter, it generates the request including this desired layer 2 destination address and forwards this to the communications restrictions module (CRM) in the network management module (NMM). This is schematically depicted by the arrow with the message "req MAC2". Upon receipt of this request, the CRM checks in the memory device M1 whether ED1 is allowed to forward packets to layer 2 destination address MAC2. This is represented by the arrow with the message schematically denoted "MAC2/ED1?" from CRM to M1 .If this address is contained within the list of M1 for device ED1, this is retrieved from M1 to CRM, and represented by the arrow back from M1 to CRM denoted "MAC2". This information is subsequently passed to the communications control device CDD of the edge node ED1 which subsequently forwards this address MAC2 to the communications restriction filter CRF, as denoted by the arrow from CCD to CRF with the message "MAC2". It is stored there for a predetermined time, for instance 120 seconds. After this time has elapsed, this entry is again removed from the filter CRF. This allows to regularly request whether layer 2 destination addresses are still allowed such as to also become regularly updated of the changes. Other possibilities for having a very dynamic mechanism comprise requesting the communications restriction module each time a new packet arrives. Then each time such a packet arrives the filter is updated, independent on whether the layer 2 destination address is already present in the filter or not. For these implementations even a filter may be omitted from the edge devices. This type of solution however requires extra communication time between the edge node and the network management module.

In the case the requested address MAC 2 is not allowed for communications from ED1, the communications restriction module CRM will thus not find this address in M1, and accordingly cannot send this address back to ED1. This address can thus also not be stored within the communication restriction filter, such that the communications control device CCD, upon receiving an incoming packet with this layer 2 address as destination, will then block this packet from entry within the access network.

In case a failure has occurred in the destination edge node ED2 of which MAC2 was a layer 2 address of one port, a new port with a new MAC address will be used. The network management module, via an ARP module therein, transfers this up-to-date information to M1, either directly or via the CRM. ARP is the abbreviation of Address Resolution Protocol, which is an existing mechanism that takes care of the distribution of the updated addresses. This is however beyond the scope of this patent, and more information about this mechanism can be found in specialised literature.

Fig. 2 shows this mechanism via an arrow denoted MAC20 between ARP and M1. The sender of the packet is also informed, using the ARP mechanism that another port having for instance layer 2 address MAC20, has to be used. M1 now contains for ED1 an updated list wherein MAC2 is omitted and instead MAC20 is added. The sender of the packet, now informed to use MAC20, inserts this new layer 2 address in the header of its packet. If CRM subsequently asks M1 whether MAC20 is allowed for ED1, M1 will provide the info that indeed this MAC20 is allowed.

Address MAC 20 is subsequently provided by CRM to CCD, which subsequently stores this information within CRF. Accordingly, if new packets will then arrive in CCD having MAC20 in their header, CCD will then get from CRF the information that MAC20 is allowed.

In addition to the retrieval of allowed layer2 destination addresses such as MAC addresses in Ethernet networks, the CRM can further get from another memory denoted VLANM in NMM additional layer 2 forwarding information such as a VLAN tag. VLAN is the abbreviation of virtual local area network and has the advantage to provide increased performance and manageability, physical topology independence and increased security. The above mentioned advantages can be used in the above described scenario for service segregation in the access network.

It is also possible that CRM gets from another memory, for instance M3 in Fig. 2, layer 3 or higher layer associated with the allowed layer 2 MAC addresses. The higher layer information is useful when certain protocols or applications have to be restricted.

CRM may be implemented as a communications restriction software agent in the network management module or as a hardware module.

In some embodiments of the method and of the edge device , the communications control device CDD of the edge node may also be able to insert, within the generated requests, user information with respect to the user which has sent the packet. This user information may comprise physical layer information such as for instance the line of the access multiplexer to which the user is connected. This user information is thus further passed through in the request to the CRM of the network management module. The latter extracts this user information from the request , and may provide this to a charging module CM. CM may accordingly respond by providing some charging information to CRM. This information is useful in case the desired layer 2 address is not stored within M1. This happens for instance for user-to-user communications within the access network itself, which are normally not allowed by the network management, since these cannot normally be charged. Using the present mechanism however, where the CRM consults the charging module CM upon receipt of such a request, the charging module becomes aware of this kind of communications, such that it can charge for it. CM can then further provide this information to CRM which can use this information to decide to allow the requested layer 2 address, thus by providing this address in response to a request including this layer 2 address from the ED1. While the principles of the invention have been described above in connection with specific apparatus, it is to be clearly understood that this description is made only by way of example and not as a limitation on the scope of the invention.