Method for the automatic configuration of communication relationships between communication units situated in a packet-oriented communications network

Within the framework of the optimization of current communication networks, in particular of broadband subscriber access networks - also referred to as access networks - a large number of subscribers are to be ensured at low cost access both to narrowband and broadband services, eg video-on-demand. In the course of the optimization, the technological and economic outlay for the realization of network facilities that can be arranged in current communication networks can be reduced by the use of technologies (eg personal computers) not developed specifically for broadband subscriber access networks but for mass markets. As an example of such a widely-spread and correspondingly developed technology, the "Ethernet" standardized according to the IEEE standard 802.3 is mentioned, which provides a frame- or packet-oriented and connectionless transmission method. In the case of network devices, such as multiplexing devices, which are arranged in current communication networks, it is known, for example according to the asynchronous transfer mode-ATM-configured data cells-also referred to as ATM cells-as well as time-slot-oriented information (eg TDM or PCM structures, pulse code modulation) Which is arranged locally in the network device, between distributed local subscriber line modules arranged in the network device and at least one central unit or subassembly having central functions. The Ethernet can be used both as a "wiring" or "backplane" in a subrack for bridging smaller distances as well as as a comprehensive communication network for bridging larger distances.

In the transition from traditional circuit-switched or time-slot-oriented communication networks to packet-oriented communication networks, in particular to communications networks according to the IEEE standard 802.3, the change of the transmission method also entails changes in the addressing and configuration of the system components involved. In current communication networks, the following system structure is often used:<ul><li>A central subassembly (hereinafter also referred to as a central communication device) with a plurality of communication units, which are provided, for example, Ethernet interfaces and are also referred to as converter modules hereinafter, which are connected to a packet-oriented communications network, eg Ethernet, via a central switching device, eg Ethernet switch ,</li><li>A plurality of decentralized modules, also referred to as decentralized communication devices, which also at least at least one communication unit -converter module -providing in each case an Ethernet interface, the at least one communication unit being connected to the communication network via the switching network and here via the Ethernet network switch Central communication device.</li></ul>

In the case of the system structure described above, the task is to enable a functional commissioning of the arrangement or the system booting, in the manner that a logical point-to-point communication between a respective decentralized communication unit or converter module associated with the central assembly is realized Point connection.

A conventional solution is to set the individual configuration of the respective point-to-point connections via the communications network by means of local management consoles which can be connected to the communication devices. However, in a spatially extended system with many remote communication devices, this is not feasible by the network operator because of the increased complexity. The alternative option of a fixed configuration is also not possible because the assignment is to be done dynamically.

Furthermore, document US-B1-6377990 discloses a system consisting of a communication network in which a gateway and a plurality of servers are located. The ARP protocol is used to determine the destination address. First, an ARP packet is sent to all stations by means of a broadcast. From the answer, the destination address is indicated accordingly. The object of the invention is to improve the configuration of communication relationships between central and decentralized communication devices arranged in a communication network in such a way that no interaction with a central management system is required. The object is achieved by a method and by a communication arrangement according to the features of the preamble of claims 1 and 9 by the respectively characterizing features.

In the method according to the invention, communication relationships are configured between communication units arranged in a packet-oriented communications network, at least one decentralized and one central communication device and each having communication network-specific address information. The essential aspect of the method according to the invention is that the at least one decentralized communication device or the at least one communication unit allocated to it at least one data packet each having the address information of the communication unit is formed and transmitted via the communication network to the central communication device. The address information contained in the received data packet is determined by the central communication device, at least one communication unit assigned to the central communication device is selected and the determined address information is assigned to the selected communication unit. Furthermore, the central communication device or the at least one selected communication unit at least forms a data packet having in each case the address information of the selected communication unit and transmits it via the communication network to the decentralized communication device. With the aid of the address information transmitted to the central and the decentralized communication device, the communication relationship between the addressed communication units is configured via the communication network.

The essential advantage of the method according to the invention is that no management system (central or decentralized) is required for setting up or for configuring the communication relationships between communication units arranged in a communication network, for example interface units or converter modules. Thus, the method according to the invention can be used during the initialization of a communication network - for example, system start-up - since communication with a central management system is still not possible at this time.

Only after setting up these communication relationships according to the method according to the invention, the management system can, for example, communicate via the communication network via the established communication relationships. The method according to the invention makes it possible, at the earliest possible time of the system initialization, to make known to the communication units or components involved in a communication connection to be set up the mutual communication network-specific address information - also referred to as MAC addresses - at runtime To-point connections.

Advantageously, the packet-oriented communication network is configured according to the IEEE standard 802.3 - claim 6. Such Ethernet networks are designed for the mass market in the vicinity of local networks - and thus cost-optimized. With the aid of the Ethernet technology, for example, internally arranged communication networks can be used locally in a network device as cost-effective wiring of, for example, central and decentralized modules - "backplane".

Further advantageous embodiments of the method according to the invention as well as a communication arrangement for carrying out the method according to the invention are to be found in the further claims.

The method according to the invention is explained in more detail below with reference to two drawings. Here<dl id="dl0001" compact="compact"><dt>FIG</dt><dd>A communication arrangement arranged in a communication network for implementing the method according to the invention</dd><dt>FIG</dt><dd>An alternative embodiment variant of the communication arrangement shown in FIG</dd></dl>

FIG. 1 shows, in a block diagram, a network device NE arranged in a subscriber access network or access network ACCESS, configured as a multiplexing device, to which, via a plurality of decentralized subassemblies DBG1... K arranged in the network device NE, and the subscriber subscriber lines, (Not shown) are connected to a higher-level communication network OKN, for example, an ISDN communications network. In the network device NE, an "internal" packet-oriented communications network EN, which is configured according to the IEEE standard 802.3 - hereinafter also referred to as "Ethernet" - is arranged, via which the respective decentralized modules DBG1... K - hereinafter referred to as decentralized Communication devices - are connected to a central module ZBG - hereinafter also referred to as a central communication device. The decentralized communication devices DBG1... K and the central communication device ZBG each have a switching device SW, adapted to the transmission technology of the communications network EN, hereinafter referred to as an Ethernet switch, wherein the communication devices DBG1.. Ethernet switch SW and a connection AD, AZ of the respective communication device DBG1 ... k, ZBG provided for this purpose are connected to the packet-oriented communication network EN.

In this exemplary embodiment, each decentralized communication device DBG1... K has a communication unit KE, which is respectively represented by a converter module KE. Each converter module KE which implements the subscriber line comprises conversion functions for the transition between packet-oriented transmission technology realized on the subscriber line, for example, and the packet-oriented transmission technology realized in the communications network, here TDM / Ethernet. Each converter module KE is connected to the Ethernet switch SW arranged in the decentralized communication device DBG1... K. It should be noted that each decentralized communication device DBG1... K can have a plurality of such converter modules or communication units KE.

Furthermore, it should be noted that, in the event that only one communication unit KE is assigned to the decentralized communication device DBG1... K, the decentralized communication device DBG1... K and the assigned communication unit can be logically combined, that is to say as an identical or logically connected unit Can be considered.

Furthermore, a plurality of converter modules or communication units KE, which are each connected to the Ethernet switch SW arranged in the central module ZBG, are also arranged in the central communication device ZBG. All communication units KE each have an address information mac1_1... K_n, macz1... 1, which uniquely identifies the respective communication unit KE in the packet-oriented communication network EN, also referred to as MAC address, medium access control. Each communication device DBG1... K, ZBG, which is connected to the packet-oriented communication network EN, has a control device STE, which is connected to the components of the respective communication device, ie to the respective converter modules KE and the Ethernet switch SW. The converter modules KE arranged in the central communication device ZBG each have a connection AL via which the respective converter modules KE are connected via corresponding outputs UL of the central communication device ZBG to the superordinate communication network OKN configured in this exemplary embodiment according to the ISDN transmission method.

According to an alternative embodiment shown in FIG. 2, the outputs AL of the converter modules KE arranged in the central communication device ZBG are fed back to corresponding connections AP 'of the Ethernet switch SW arranged in the central module ZBG and connected to the superordinate communications network via further connections AP (KE) and the superordinate communications network (OKN) are transmitted via the Ethernet switch SW and its connections AP ', AP ", respectively.

In the course of the commissioning of the arrangement shown in the block diagram, communication relations or logical point-to-point connections between the communication units KE arranged in the decentralized and the central communication devices are now to be set up via which, after successful commissioning, useful information is transmitted.

The sequence of the method according to the invention is explained in more detail below: <ol><li>1. In the case of an embodiment variant of the method according to the invention, a freely selectable MAC address iadr - hereinafter also referred to as an initialization address - is defined as all control units STE which are arranged in the respective decentralized and arranged in the central communication device DBG1 ... k, ZBG. In the following, the time of the initialization of the communication arrangement shown in the block diagram - also referred to as "ramping" or "booting" of the system - is considered. After the initialization of the system or of the communication arrangement, all communication devices DBG1... K, ZBG connected to the packet-oriented communication network EN behave passively. For the exemplary embodiment, it is also assumed that, in the context of the initialization of the system, a communication relationship or logical point-to-point connection kb, which is represented by a dashed double arrow in the block diagram for many, between the converter module KE arranged in the first communication device DBG1 A converter module KE arranged in the central communication device ZBG should be set up.</li><li>2. The control device STE, which is arranged in the first decentralized communication device DBG1, configures the initialization address iadr as a (temporary) destination MAC address in the communication unit KE assigned to it. Furthermore, this initialization address iadr is configured in the manner in the assigned Ethernet switch that the connection port AP implementing the connection to the communications network EN is selected by the initialization address iadr on the network side. Furthermore, all units arranged in the communication network EN, such as, for example, switches, are configured by default in such a way that the data packets or Ethernet frames exhibiting iadr are mediated in the direction of the central communication device.</li><li>3. The control unit STE, which is arranged in the central communication device ZBG, configures the initialization address iadr in the associated Ethernet switch SW as belonging to exactly one output port AP, which is assigned exactly to one converter module KE, so that a clear switching decision immediately without "flooding" and "learning "Can be felled. According to the present exemplary embodiment, two design variants are possible: on the one hand, an incoming data packet or Ethernet frame containing the initialization address iadr as the destination address is conveyed via the converter module KE and via an interface to the control unit STE provided for this purpose-illustrated by the path L1, L2 in the block diagram . On the other hand, an Ethernet frame which has the initialization address iadr can be switched directly from the Ethernet switch SW to the control unit STE via a dedicated interface for this purpose, illustrated by the path L3 in the block diagram.</li><li>4. After the initialization according to step 2, the converter module KE of the first decentralized communication device DBG1 cyclically starts to send special data packets or Ethernet frames - in the block diagram, schematically represented by a data packet fr_z - to the initialization address iadr. These Ethernet frames fr.sub.z can, for example, be designed in such a way that they differ in some way from a normal payload and are thus clearly recognizable as a "handshake". This can, for example, be achieved by setting a specific value of the "Ethernet type field" defined in the Ethernet standard. This value is freely selectable, apart from the values ​​already provided or reserved in the standard. Furthermore, the MAC address - here mac1_1 - of the converter module KE is inserted as a "sender address" into the Ethernet frames fr_z.</li><li>5. The Ethernet switch SW, arranged in the first decentralized communication device DBG1, forwards the data packet fr_z formed in this way via the port AP to the Ethernet EN since this switching decision is unambiguous by step 2.</li><li>6. The data packet fr_z transmitted via the communication network EN to the central module ZBG is received by the Ethernet switch SW arranged in the central communication device ZBG, the sender address mac1_1 contained in the received data packet fr_z being determined or "learned" Switch SW is also made a clear switching decision based on the configuration described in step 3. The received data packet is relayed via one of the configured port ports AP to the control device STE of the central communication device ZBG either via the illustrated path L1, L2 or through the alternative path L3, depending on the selected embodiment.</li><li>7. The control device STE, which is arranged in the central communication device ZBG, extracts the sender MAC address (here mac1_1) contained therein from the received data packet or Ethernet frame fr_z and selects one of the currently available and unassigned ones in the central communication device ZBG Converter module KE. For this embodiment, it is assumed that the transformer module KE with the associated MAC is selected maczl address. The extracted sender MAC address (mac1_1) is implemented by the control device STE in the selected converter module KE (macz1) as the target for the sender direction of the point-to-point connection to be established. Furthermore, the MAC address of the selected converter module (here maczl) is determined by the control device STE, which must be communicated as a (final) destination to the first decentralized communication device DBG1.</li><li>8. The control device STE arranged in the central communication device ZBG forms a similarly "special" Ethernet frame (for example via the path L3 shown in the block diagram) in response to the Ethernet frame transmitted by the decentralized communication device DBG1. The first decentralized communication device DBG1 or the communication unit or converter module KE (mac1_1), which is assigned to this decentralized communication device DBG1, ie the MAC address mac1_1, which has just been learned, is the goal of this particular Ethernet frame or data packet-schematically represented in the block diagram by a data packet fr_d Is inserted into the Ethernet frame fr_d as the target information. Furthermore, the sender MAC address of the converter module selected in step 7 (here maczl) is inserted as the sender address for the special Ethernet frame fr_d. All the Ethernet switches SW involved have learned the MAC address of the converter module KE (mac1_1) assigned to the first decentralized communication device DBG1 and can make a clear switching decision for the return path.</li><li>9. The Ethernet switch SW arranged in the first decentralized communication device DBG1 extracts from this received data packet or response Ethernet frame fr_d analogously to step 3. Thus, the control device STE can extract the sender MAC address contained in the Ethernet frame and As the final destination of the point-to-point connection kb to be set or configured, in the converter module KE assigned to it.</li><li>10. The converter module KE arranged in the first decentralized communication device DGB1 stops the cyclic transmission of Ethernet frames fr_z according to step 4, so that the transport of the actual payload can now begin via the communication relationship kb set up with the aid of the method according to the invention.</li></ol>

According to an alternative embodiment variant of the communication arrangement shown in block diagram, the respective decentralized communication devices DBG1... K can each be connected to the central communication device ZBG via a connecting line implementing the communication network EN - each illustrated by a dashed connecting line VL in the block diagram. Via these connection lines VL, all data packets transmitted by the decentralized communication devices DBG1... K are automatically transmitted or routed to the central communication device ZBG. It should be noted that in such an arrangement, the random MAC address or initialization address iadr, which has hitherto been provided for the method according to the invention, is not required, since all data packets without switching decisions are transmitted via the connection lines from the decentralized communication devices DBG1 Communication device ZBG, and thus the initialization address as the destination information or routing information for the Ethernet frames is not required. In this embodiment variant, therefore, only the MAC addresses of the respective converter modules KE, which represent the end points of the point-to-point connections to be set up, ie, mac1_1... K_n, macz1... 1, must be inserted and transmitted in the described manner. However, the non-use of routing information for Ethernet frames would not be standardized (eg according to IEEE 802.3) so that the method according to the invention would not be feasible with standard-compliant components. A proprietary adaptation associated with greater effort would be required, although any type of management (centralized or decentralized) can be saved in this design variant.

The method according to the invention makes it possible to dynamically configure and system-configure a system configuration without the participation of a central management system. The method shown in more detail in the exemplary embodiment uses only means according to the Ethernet standard and comes without broadcasts which are regarded as unsuitable, particularly in the environment of communication networks operated by different network operators, for reasons of clear data separation. With the exception of the embodiment variant described above, the method according to the invention requires a single fixed but freely selectable MAC address or initialization address in order to be able to configure a basically unlimited number of point-to-point connections.