WO2007124859A1

WO2007124859A1 - Configuration of devices of a data transfer network using identification information of a connecting unit

Info

Publication number: WO2007124859A1
Application number: PCT/EP2007/003356
Authority: WO
Inventors: Magnus Damberg; Peter Sandberg; Linus Thrybom
Original assignee: Bombardier Transportation Gmbh
Priority date: 2006-04-27
Filing date: 2007-04-11
Publication date: 2007-11-08
Also published as: GB2437580A; GB0608509D0

Abstract

The invention relates to a method of configuring a data transfer network (1) for transferring data to and/or from a plurality of devices (4a, 4b), wherein the network (1) comprises at least one configuration server (11) and a connecting unit (3) for connecting at least one of the devices (4a, 4b) to other parts of the network (1) and wherein the method comprises the following steps: (a) A first message is transferred from one of the devices (4a, 4b) to the server (11) via the connecting unit (3) wherein the first message comprises a name of the device (4a, 4b; 14a, 14b, 14c, 14d); (b) the connecting unit (3) adds identification information to the first message, wherein the identification information identifies the connecting unit (3) and/or a port (A, B) which the device (4a, 4b) is connected to; (c) the configuration server assigns at least one configuration parameter to the device (4a, 4b), depending on the device name and on the identification information.

Description

CONFIGURATION OF DEVICES OF A DATA TRANSFER NETWORK USING IDENTIFICATION INFORMATION OF A CONNECTING UNIT

The invention relates to the configuration of a data transfer network for transferring data to and/or from a plurality of devices, wherein the network comprises at least one configuration server and a connecting unit for connecting at least one of the devices to other parts of the network.

In particular, this invention relates to a method for configuring data transfer networks in railway trains. However, the invention can also be applied to data transfer networks for industrial purposes, for example.

According to the OSI reference model (as described for example by Wikipedia, the free encyclopaedia, in the Internet, http://en.wikipedia.org/wiki/OSI_model), network communications are divided into a plurality of protocols within layers of the model. Local Area Networks (LANs) operate using medium access protocols within lower layers (layers 1 and 2) of the OSI model, such as the carrier sense multiple access with collision detection CSMA/CD (IEEE Standard 802.3), also known as ETHERNET, and the token ring access ring method of IEEE Standard 802.5. These two lower layers can be split into the physical layer and the data link layer, wherein the data link comprises a media access control (MAC) layer and a logical link layer.

Systems, such as personal computers, workstations and mainframe computers, which can be connected to the LANs, each have a distinct lower level protocol identifier known as the physical network address or MAC address. LAN data (frames) transferred to a destination system on the network under these lower level protocols contain the destination system MAC address, or other physical network address, as a destination address. However, systems communicate within the data transfer network by encapsulating additional protocols (higher OSI layers 3-7) within the lower layer LAN frames. These higher level protocols are grouped into suites such as the TCP/IP protocol suite and the XNS protocol suite. Many LANs contain groups of end systems (terminal devices) that use different higher level protocol suites. These higher level protocol suites also assign unique higher level protocol identifiers to systems which transmit or receive frames in the network.

For instance, an internet protocol IP address is assigned to each device or system operating within an internet protocol network. The internet protocol address includes a network address portion and a host address portion. The network address portion identifies a network within which the system resides, and the host address portion uniquely identifies the device or system in that network. Processors routing packets in an internet protocol network rely on the network address portion of the IP address in a frame to find the local network of the destination device. Once the local network of the destination is located, the frame is forwarded to that network where the host address portion is relied upon to assign a MAC address for the destination device to the packet. Thus, the higher level protocol can effectively manage the routing of packets among the networks, without maintaining a table of the unique physical access layer identifiers for all of the terminals in the network.

In order to communicate in such a network, the device must first obtain its higher level protocol address, which can be assigned by a central authority, such as the Internet Activities Board, or by a network manager. Normally, a particular device learns its IP address by a configure operation, in which a technician uses a local terminal to configure the device. In a centrally managed network, this could be a cumbersome task, involving travel of skilled personnel away from the central management location.

US2002/0046263 A1 describes a method of configuring an automation module on a TCP/IP network or on a TCP/IP sub-network, to which at least one piece of automation equipment is also connected. The term "automation module" denotes any equipment provided with a processing unit and a communication interface with a TCP/IP network and used in an automation assembly, for example like an in- puts/outputs module, a speed controller, a regulation device, a man/machine dialog terminal, a programmable logic controller or any specific module of a programmable logic controller or a numerical control.

The configuration method according to US 2002/0046263 A1 comprises the following steps in sequence: a preliminary step to assign a single application name on the TCP/IP network for the automation module; an addressing step in which the automation module sends a request address query on the TCP/IP network containing the application name of the automation module and conform with the DHCP protocol, to a DHCP server; a configuration step in which the automation module sends a read configuration query to the FTP/TFTP server conform with the FTP or TFTP protocol, on the TCP/IP network. During the addressing step, the automation module receives a response to the request address query from the DHCP query server, containing an IP address and the location of a data file specific to the automation module, making it possible to go on to the configuration step. During the configuration step, the automation module receives a response to the read configuration query from the FTP/TFTP server, containing the automation module data file that can then change to an operational state.

According to Fig. 4 and the related description of the document, an application name is assigned in a preliminary step. The purpose of this step is to assign to the automation module 10 an application name 40 that must be unique in the TCP/IP network or sub-network accessible to the DHCP server. This application name is composed of an ASCII character string. The assignment step may indifferently be made before the module is connected to the TCP/IP network. Many assignment means to assign an application name to a module can be used; local assignment using a Man-Machine interface, assignment by local communication (serial link, etc.) or remote communication (web page), assignment by mechanical means (encoder wheels, switches, etc.). A combination of these different assignment methods is also possible. Furthermore in some modules the application name can also be formed by concatenation of characters predefined by the manufacturer to characterize a type of family of modules, in combination with a customisation number assigned by one of the assignment methods mentioned. Once determined, the processing unit of the automation module is capable of memorizing the application name in the storage means. When a defective automation module is replaced by a new module, the application name used for the defective module will be reassigned to this new module so that it can retrieve the data file for the defective module.

In a following addressing step, the automation module must be connected to the TCP/IP network accessible to a DHCP server. The DHCP client for the automation module can then send a request address query conform with the DHCP protocol to the DHCP server installed in the automation equipment. This DHCP request address query contains an optional DHCP field (client identifier code) containing the application name that was memorized in the storage means of the automation module, as a parameter. When the DHCP request address query is received, the DHCP server searches in a configuration table memorized in storage means of the automation equipment for an IP addressing and a location of the data file corresponding to the received application name.

Any application name of the automation module must firstly be input by appropriate means into the configuration table of a DHCP server by associating an IP addressing with the automation module. Once the configuration table has been updated, a module can be replaced by another module with the same application name using the method described in this document.

While the method according to US 2002/0046263 A1 makes it possible to replace a defective device, the effort for administrating the configuration information is high. In particular, at least some reprogramming is required. A computer system administrator is needed in order to perform this task. Furthermore, it is not possible to swap the physical locations of devices, wherein these devices provide identical or similar functionality.

There are other ways of configuring terminal devices in data transfer networks. According to one configuration method, the network server and the terminal device can communicate using broadcast telegrams in order to obtain and assign an IP-address to the terminal device. However, this requires low level access to the communications tack and the effort is high. According to a second solution, the IP-address of the terminal device can be received by the terminal device using standard DHCP functionality. After identification of the terminal device, the server sends the pre-configured IP-address to an application running in the terminal device, which then sets the address. However, the last step, the setting of the address, does not comply with standard DHCP.

According to a third approach, the terminal device again receives the IP-address using standard DHCP functionality. After identification of the terminal device, the server uses SNMP (simple network management protocol) to set the pre-configured IP- address in the terminal device. A disadvantage of this procedure is that a SNMP client would be required in every terminal device, wherein the SNMP client must be capable of setting the IP-address.

According to a forth way of configuring, the terminal device receives an IP-address using standard DHCP, but with a short lease time. After identification of the terminal device by the server, the lease will expire and the DHCP-server then assigns a new (pre-configured) IP-address to the terminal device. However, this means a lengthy configuration process, since the lease time cannot be set to short values, because the process would not be reliable otherwise. Furthermore, short lease times may not be supported by standard DHCP clients.

According to a fifth method of configuration, the terminal device receives an IP- address using standard DHCP functionality. After identification of the terminal device, the server sends on application level a request to the terminal device in order to renew the lease. The DHCP-server then assigns a new (pre-configured) IP-address to the terminal device in order to replace the previously given address. Again, the configuration process would be rather lengthy and, in addition, this would require an application in the terminal device that initiates the DHCP-renewal procedure.

It is an object of the present invention to provide a configuration method which is easy to perform, which is fast, which reduces the load of traffic within the network and which makes it possible to exchange or replace devices of similar or identical type. Preferably, only standard procedures for the configuration of data transfer networks should be used. The following is proposed: A method of configuring a data transfer network for transferring data to and/or from a plurality of devices, wherein the network comprises at least one configuration server and comprises a connecting unit, in particular a network switch, for connecting at least one of the devices to other parts of the network and wherein the method comprises the following steps:

• a first message is transferred from one of the devices to the server via the connecting unit, wherein the first message comprises a name of the device,

• the connecting unit adds identification information to the first message, wherein the identification information identifies the connecting unit and/or a port which the device is connected to,

• the configuration server assigns at least one configuration parameter to the device, depending on the device name and on the identification information.

"Adding identification information to the first message" includes the case that the connecting unit creates and transfers a new message which comprises the first message and, additionally, the identification information. Consequently, the invention includes both options: the first message is amended by adding the identification information and/or the new message is created. If the configurations server receives both the first message (without the identification information) and the new message (containing the identification information), it is preferred that the configuration server disregards the message which does not comprise the identification information, i.e. the first message. Furthermore, it is preferred that the message which contains the identification information is sent from the connecting unit to the configuration server using uni- cast (in contrast to broadcast, which would increase the traffic unnecessarily).

In particular, the first message can be a standard DHCP request message, which is a broadcast message. The DHCP server is configured to disregard all request messages which are sent as broadcast messages. The connecting unit can therefore modify the first message or can create the new message, wherein the modified first message or the new message is sent as unicast message to the DHCP server. Consequently, the DHCP server will accept the unicast message. The data transfer network may be, for example, one of the networks mentioned above, in particular a network in a railway train. Typically, there is a plurality of connecting units in a data transfer network of a railway train. For example, a railway car comprises a plurality of network switches, wherein in each case a single door controller for controlling the opening and closing of a door is connected to the network switches. Furthermore, the network in the train may comprise redundant servers.

In particular, the device from which the first message is transferred may be a terminal device, i.e. a device which can be a source or sink of data transferred via the network during normal operation of the network. Connecting units like switches or hubs are not terminal devices. However, a local network (e.g. a LAN) is considered to be included by "terminal device".

Preferably, the configuration server is a DHCP (Dynamic Host Configured Protocol) server. The DHCP provides a framework for passing configuration information to hosts on a TCP/IP network. A documentation of the DHCP can be found in the Internet, http://rfc.net (document rfc2131 ).

The connecting unit may be any kind of device, which is capable of connecting at least one data transfer line of the network to one or more than one other data transfer lines. In particular, the connecting unit may define a node where a plurality of data transfer lines of a sub-network is connected together. For example, a star topology has such a node. Preferably, the connecting unit is a network switch. Such a device is often referred to as "intelligent hub".

US 2002/0046263 A1 discloses (as mentioned above) a method for configuring network devices, wherein terminal devices have names and configuration parameters such as an IP address, but also a sub-network mask and a gateway address, for example, are selected to the device by the server on the basis of a configuration table, wherein the configuration parameters are assigned to the name in the table. According to one embodiment of the present invention, the following steps can be performed as described in US 2002/0046263 and the relevant disclosure of the document (in particular the text on page 2, beginning of right column to page 3, end of left column) is herewith incorporated by reference: a) the assignment of a name to the device; b) the assignment of an address to the device; and c) the configuration of the device.

However, there are the following essential differences between the method according to US 2002/0046263 and the method according to the present invention:

The connecting unit adds the identification information to the message which is sent from the device to the server and this identification information is used (in addition to the name) to find the at least one configuration parameter (e.g. the IP address). This facilitates the handling and preparation of the configuration process and will reduce the load within the network during the configuration process and will accelerate any modification of the configuration, when devices are exchanged or replaced. These advantages will become apparent from the following.

A standard DHCP client can be used in the device for generating the first message.

The same name or the same part (e.g. type name) of the name may occur more than once in the data transfer network. In this case, the name may comprise a type name, wherein the type name identifies a type of the device. Thus, the combination of the type name (which identifies the type of the device) and the identification information (which identifies the connecting unit) can uniquely identify a particular location of the device, although further devices of the same type may exist within the network. Consequently, two devices of the same type can be exchanged with each other (i.e. each of the devices is connected to the network at the location of the other device) and it is not necessary to update the configuration information (such as a configuration table) which is used by the server in order to find the configuration parameters. As long as there is only one item of the same device type connected to the connecting unit, the device name can be identical to the type name. There is no need to use names which individualise the devices further.

However, according to another embodiment, the identification information may comprise a plurality of devices of the same type which are connected to the same connecting unit. In this case, the name can also comprise additional information (e.g. a number which uniquely defines the corresponding device, at least within the local network of the connecting unit) and/or the identification information may comprise information about the port which a particular terminal device or sub-network is connected to. In this embodiment, the identification information also facilitates the configuration, since it comprises additional information in order to identify the location and/or function of the device. For example, there may be two devices of the same type connected to the same connecting unit. If only one of the two devices is moved to another area of the network or is replaced by a new device, the server can determine this fact from the identification information and from the name of the device. Care must be taken only, if both of the two devices are moved or exchanged.

More generally speaking, the name may consist of the type name (according to one embodiment) or may comprise the type name and additional information in order to distinguish the device from other devices of the same type (according to another embodiment).

The term "type" is understood to define a class of devices which perform similar or identical functions and can be replaced by each other. This means that different devices of the same type may have different construction and different functionality, but are capable of performing the same task (as defined for a particular application).

Preferably, the configuration server is a DHCP (Dynamic Host Configured Protocol) server, wherein the first message is a DHCP discover message and/or a DHCP request address query, as described in US 2002/0046263 A1 for example. Consequently, the method according to the present invention can be applied repeatedly. For example, the method can be applied for a first time, when a DHCP discover message is sent from the client to the server via the connecting unit, and it can be applied for a second time during the same configuration process, when a DHCP request message is sent from the client to the server via the connecting unit. In addition or alternatively, the connecting unit can send the message, which comprises the identification information, to more than one server, for example to redundant servers (i.e. servers which have the same functionality with respect to the configuration procedure). Redundant servers are used to increase overall system availability. Furthermore, it is proposed that the identification information is added by the connecting unit to the first message using the Relay Agent Information Option of the DHCP. This includes the above-mentioned case of creating a new message which contains information of the first message and the identification information. A documentation of this option can be found in the Internet, http://rfc.net (document rfc3046).

According to another aspect of the present invention, the following is proposed: A data transfer network for transferring data to and/or from a plurality of devices, wherein the network comprises at least one configuration server and at least one connecting unit for connecting at least one of the devices to other parts of the network and wherein:

• at least one of the devices connected to the connecting unit is adapted to send a first message to the server via the connecting unit, wherein the first message comprises a name of the device,

• the connecting unit is adapted to add identification information to the first message, wherein the identification information identifies the connecting unit and/or a port which the device is connected to,

• the configuration server is adapted to assign at least one configuration parameter to the device, depending on the device name and on the identification information.

It is possible that the connecting unit and the server are integrated in the same physical device. For example, the connecting unit may be a network switch which also comprises the server (or at least one of redundant servers).

A railway vehicle or railway train may comprise the data transfer network or part of the data transfer network. For example, the railway vehicle (such as a car of a train) may comprise at least the device and the connecting unit, but the server may be located in a different part of the train. Examples and preferred embodiments of the invention will be described in more detail in the following with reference to the accompanying drawing. The figures of the drawing show:

Fig. 1 a data transfer network, in particular located within a railway train;

Fig. 2 a flow chart illustrating steps of a process of configuring terminal devices of the network shown in Fig. 1 ;

Fig. 3 schematically a railway car comprising a part of a data transfer network for the control of four doors;

Fig. 4 the data format of a message for transferring the identification information; and

Fig. 5 the detailed structure of a part of the message shown in Fig. 4.

The network 1 shown in Fig. 1 comprises a network control unit 12 having a DHCP- server 11. Optionally, there may be additional servers in the network 1 , such as the server 13, which is a redundant server located at the same or different area of the network 1. In addition or alternatively, an additional server having a different functionality compared to the server 11 may be part of the network 1. In this case, controlling the operation of the devices, which are connected to the network, can be performed by a different server than the configuration of the devices, for example.

A plurality of devices is connected to the network 1. Only two of the devices, marked with reference numerals 4a, 4b are shown in Fig. 1. The device 4a is connected to a port A of a network switch 3 via line 16. The device 4b is connected to port C of the network switch 3 via line 17. The device 4b may be a controller for controlling the operation of other equipment such as a door 7 of a railway vehicle. The small rectangles within the devices 4a, 4b represent client software. Each of the devices 4a, 4b preferably comprises a standard DHCP client.

Reference numeral 15 represents other areas of the network 1 which are not shown in more detail in Fig. 1.

Pre-defined configuration information, such as a configuration table, may be stored locally at the control unit 12 and/or separately from the control unit 12. Reference numeral 9 denotes a storage unit containing the pre-defined configuration information. For example, the storage unit 9 may be removable from the network 1 and may be plugged into a connector before the configuration process.

Fig. 2 illustrates a preferred embodiment of the configuration method of the present invention. Three vertical lines are shown in Fig. 2. The left line is marked with reference numeral 4b and represents the corresponding terminal device, for example device 4b of Fig. 1. The vertical line in the middle is marked with "3" and denotes a connecting unit, such as the network switch 3 of Fig. 1. The vertical line on the right hand side is marked with "11" and represents the configuration server, for example the DHCP-server of Fig. 1.

Before the configuration process starts, the configuration information is prepared and a name is assigned to each terminal device or other device in the network which is to be configured according to the method of the present invention. This assignment is made as part of the process of preparing the configuration information and, as well, the name is assigned to the respective physical device, such as the device 4b. However, the name which is assigned to the physical device may differ from the name used in the consideration information by additional information in order to uniquely identify a specific physical device.

The name may be a name characterizing a type of the device, for example "door- controller". The configuration information is prepared by logically assigning IP- addresses to devices, which are expected and/or required for a specific application (like controlling the operation of a railway car). For example, four door-controllers may be expected, two for opposite doors at one end of the car and two for opposite doors at the other end of the car, for each car of a railway train. Furthermore, a particular structure of the data transfer network can be considered for the configuration information, such as the fact that each of the door controllers is connected to a separate connecting unit (in this case it is not necessary that the name comprises additional information in order to uniquely identify a specific physical device).

Fig. 3 shows such a structure. A railway car 20 is connected to the data transfer network of a railway train. As indicated by reference numeral 15, this area of the network can be connected to the network 1 according to Fig. 1. At the front end of the car 20 (at the top of Fig. 3) the network 15 branches off to a first network switch 3a and to a second network switch 3b. In practice, a further network device, such as a switch or router, may be located at the branching point(s) of the network. However, these further network devices are omitted in Fig. 3. A first door-controller 14a, which is a terminal device of network 15, is connected to a port of network switch 3a. The door- controller 14a is adapted to control the operation of a first door 7a of car 20. In corresponding manner, a second door 7b is controlled by a second door-controller 14b which is connected to a port of the second network switch 3b. A similar arrangement of two network switches 3c, 3d, door-controllers 14c, 14d and controlled doors 7c, 7d is located at the end of car 20 (at the bottom of Fig. 3).

Further devices (not shown in Fig. 3) may be connected to the network switches 3a to 3d.

The configuration information contains information about the fact that there should be four door-controllers in car 20, wherein each of the door-controllers 14a to 14d is connected to a separate network switch 3a to 3d. Thus, it will be sufficient for the configuration server to receive the information that a door-controller which is connected to a particular switch requests the assignment of an IP-address. With this information, the configuration server can determine the corresponding IP-address which is logically assigned to the door-controller within the configuration information.

The preferred embodiment of setting an IP-address to a device in a data transfer network will be described in the following with reference to Fig. 2.

In a first step SO the client, which is installed in device 4b, receives and/or retrieves the device name, which preferably consists of the device type. For example, the device name can be stored in a data storage within the device 4b or can be typed in by an operator. The client generates a DHCP discover message, which comprises the device name (host name). The corresponding DHCP option is described in document RFC1533 (http://rfc.net/rfc1533.html), Section 3.14. Host Name Option. This document defines the format of information in the last field of DHCP packets (Options'). The code of the option is "12". This option is therefore often referred to as "option 12".

This (first) message is transferred to the connecting unit 3 as indicated by the arrow in the top left corner of Fig. 2.

In step S1 , the connecting unit 3 adds identification information to the received message, according to the DHCP relay agent information option (see document RFC3046). At least, the identification information comprises information about the identity of the connecting unit 3, such as an MAC-address and/or an IP-address of the connecting unit 3. In cases, where more than one device of the same type may be connected to the same connecting unit, the identification information preferably comprises also the number of the port which the terminal device 4b is connected to. This will enable the configuration server to detect a change of port or the fact that a device is removed from a port and transferred to another port.

The message modified in step S1 (which may also be a new message containing information from the first message and further containing the identification information) is then transferred to the configuration server 11. The configuration server 11 will now determine the IP-address or other configuration parameters which are to be assigned to the terminal device 4b and will generate a second message, in particular a DHCP offer message (step S2) which will be sent to the connecting unit 3.

Optionally, the connecting unit 3 removes information from the second message. For example, this information may be used to configure the operation of the connecting unit 3 with respect to the data transfer between the terminal device 4b and the connecting unit 3. This optional removal of information is performed in step S3.

Then, the second message is transferred from the connecting unit 3 to the terminal device 4b and, in step S4, the client of the terminal device 4b reads the configuration parameters and performs the configuration in a corresponding manner. For example, the configuration parameters may comprise the IP-address which has been assigned to the terminal device by the configuration server and the client will set the IP- address for operation within the network. A specific example of step S1 is described in the following. In the specific example, the connecting unit 4 is a network switch and the network switch is adding the information about its own MAC address and about the port (which the client is connected to) to the received first message. In particular, adding this information is performed by generating a new unicast message to be sent to the DHCP server (or to the redundant servers).

The new message is generated by the network switch according to DHCP Relay Agent Information option. Document RFC3046 states that the Relay Agent Information option is organized as a single DHCP option that contains one or more "sub- options" that convey information known by the relay agent. Sub-options are defined in particular for a relay agent that is located in specific area within the network. The sub-options include a so-called "circuit ID for the incoming circuit", and a so-called "remote ID" which provides a trusted identifier for the remote high-speed modem. In the example described here, the connecting unit is the relay agent. The MAC address of the connecting unit is transferred as the "remote ID" and number of the port, which the client is connected to, is transferred as the "circuit ID for the incoming circuit".

The data format of the unicast message of this example is shown in Fig. 4 and 5. However, this message may be part of a longer message which may also comprise additional information like a message header and other messages. Nevertheless, it is only referred to "the message" in the following.

The data format of the whole message of the relay agent information option is shown in Fig. 4. It comprises a first data field denoted by "C" for "code" which contains the value "82" for the relay agent information option. The next data field denoted by "L" for "length" contains the number of bytes in the remainder AIF (the agent information field) of the message. Here, the remainder AIF contains data of two sub-options i1 , \2.

The data structure of the whole remainder AIF is shown in Fig. 5, wherein each of the squares shown in the figure corresponds to one byte of the remainder (in the order from left to right). The data of the first sub-option i1 contains four data field, namely a first data field denoted by "SO" for "sub-option code" (here "1"), a second data field denoted by "L" for "length" (here "2", indicating that only two bytes follow within the data of the first sub-option M ), a third and a fourth data field commonly denoted by "ACID" (agent circuit ID). Here, the value of the ACID is "03" which means that the client is connected to port "3" of the network switch.

The data of the second sub-option i2 contains eight data field, namely a first data field denoted by "SO" for "sub-option code" (here "2"), a second data field denoted by "L" for "length" (here "6", indicating that six bytes follow within the data of the second sub-option i2), and six consecutive data fields commonly denoted by "ARID" (agent remote ID). Here, the value of the ARID is "10:20:30:40:50:60" which means that the network switch has the MAC address "10:20:30:40:50:60".

According to a preferred embodiment of the present invention, the server is (or the redundant servers are) adapted to interpret the "circuit ID for the incoming circuit" as the port number of the client ("the device") and the "remote ID" as the MAC address of the connecting unit. For example, the server(s) comprise(s) corresponding software for this interpretation.

In addition or alternatively, the identification information (in particular the MAC address) can comprise information which enables the server to identify the manufacturer and/or the type of the connecting unit. In this case, the server may identify the manufacturer and/or the type of the connecting unit and may use this information for making the decision how the identification information is to be used. For example, there may be a look-up table implemented in software and/or hardware within the server and the server may use information within the look-up table for making the decision. The decision may be that the "circuit ID for the incoming circuit" is interpreted as the port number of the client ("the device") and the "remote ID" is interpreted as the MAC address of the connecting unit. Consequently, the assignment of an IP address may be offered to the device (the client) on the basis of this information by the server, which may send a corresponding DHCP offer message.

Claims

1. A method of configuring a data transfer network (1 ) for transferring data to and/or from a plurality of devices (4a, 4b, 12, 13; 14a, 14b, 14c, 14d), wherein the network (1 ) comprises at least one configuration server (11 ) and a connecting unit (3; 3a, 3b, 3c, 3d) for connecting at least one of the devices (4a, 4b; 14a, 14b, 14c, 14d) to other parts of the network (1 ) and wherein the method comprises the following steps:

• a first message is transferred from one of the devices (4a, 4b; 14a, 14b, 14c, 14d) to the server (11 ) via the connecting unit (3; 3a, 3b, 3c, 3d), wherein the first message comprises a name of the device (4a, 4b; 14a, 14b, 14c, 14d),

• the connecting unit (3; 3a, 3b, 3c, 3d) adds identification information to the first message, wherein the identification information identifies the connecting unit (3; 3a, 3b, 3c, 3d) and/or a port (A, B) which the device (4a, 4b; 14a, 14b, 14c, 14d) is connected to,

• the configuration server assigns at least one configuration parameter to the device (4a, 4b; 14a, 14b, 14c, 14d), depending on the device name and on the identification information.

2. The method of claim 1 , wherein the configuration server is a DHCP (Dynamic Host Configured Protocol) server and wherein the first message is a DHCP discover message.

3. The method of claim 1 or 2, wherein the configuration server is a DHCP (Dynamic Host Configured Protocol) server and wherein the identification information is added to the first message using the Relay Agent Information Option of the DHCP.

4. The method of one of the preceding claims, wherein the name comprises a type name, wherein the type name identifies a type of the device (4a, 4b; 14a, 14b, 14c, 14d).

5. The method of one of the preceding claims, wherein the network comprises a plurality of devices (14a, 14b, 14c, 14d) of the same type and wherein only one of the devices (14a, 14b, 14c, 14d) of the same type is connected to the server via the same connecting unit (3a, 3b, 3c, 3d).

6. A data transfer network for transferring data to and/or from a plurality of devices (4a, 4b, 12, 13; 14a, 14b, 14c, 14d), wherein the network comprises at least one configuration server (11 ) and a connecting unit (3; 3a, 3b, 3c, 3d) for connecting at least one of the devices (4a, 4b; 14a, 14b, 14c, 14d) to other parts of the network and wherein:

• at least one of the devices (4a, 4b; 14a, 14b, 14c, 14d) connected to the connecting unit (3; 3a, 3b, 3c, 3d) is adapted to send a first message to the server via the connecting unit (3; 3a, 3b, 3c, 3d), wherein the first message comprises a name of the device (4a, 4b; 14a, 14b, 14c, 14d),

• the connecting unit (3; 3a, 3b, 3c, 3d) is adapted to add identification information to the first message, wherein the identification information identifies the connecting unit (3; 3a, 3b, 3c, 3d) and/or a port (A, B) which the device (4a, 4b; 14a, 14b, 14c, 14d) is connected to,

• the configuration server (11 ) is adapted to assign at least one configuration parameter to the device (4a, 4b; 14a, 14b, 14c, 14d), depending on the device name and on the identification information.

7. A railway vehicle (20) or railway train, comprising a data transfer network or part of the data transfer network of the preceding claim.