EP2837163A1

EP2837163A1 - Method for address initialization in communications systems

Info

Publication number: EP2837163A1
Application number: EP13704603.3A
Authority: EP
Inventors: Volker Blaschke; Timo Lothspeich; Andreas Mueller
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2012-04-12
Filing date: 2013-02-14
Publication date: 2015-02-18
Also published as: DE102012205991A1; WO2013152884A1

Abstract

The invention relates to a method for address initialization in a communications system (1) comprising a control entity (10) and at least one network element (101, 102, 103, 110). The network element has an identifier (1012, 1022, Δt₁₀₁, Δt₁₀₂). The method comprises the sending of a message (210, 213, 410, 601, 604, 801, 804) from the network element to the control entity, the message being sent at a time (Τ₁₀₁, Τ₁₀₂) dependent on the identifier for the network element. The method further comprises the assignment of an address (1011, 1021, 1031) to the network element. The invention also relates to a communications system which is set up to execute the method, to a computer program comprising program code means for executing method steps and to a machine-readable storage medium comprising a computer program of this type.

Description

description

title

Method for address initialization in communication systems

The present invention relates to a method for address initialization in a communication system. Furthermore, the invention relates to a communication system, a computer program for carrying out steps of the method and one or more machine-readable storage media, on which instructions for carrying out method steps are stored.

State of the art

In many communication systems, several nodes can communicate with each other via a common communication medium. The OSI

According to the layer model, the functions of the communication are divided into seven superimposed abstraction layers. In particular, the so-called link layer (also referred to as "layer 2") regulates the access of the nodes to the communication medium A particular node of such a communication system can usually be explicitly addressed by another node, which is necessary, inter alia, if the access is In this case, each device connected to the communication medium typically has a so-called layer 2 address.

In order to be able to realize a unique addressing on layer 2, it is important that the layer 2 addresses used are unique at least on the link on which the respective addresses are used.

Frequently used are the so-called EUI-48 or EUI-64 addresses managed by the Institute of Electrical and Electronics Engineers (IEEE), which are intended to be globally unambiguous. for example during a production onsschritts) in the device, without thereby generating a potential address conflict.

However, global uniqueness requires that the addresses themselves be correspondingly long: For example, each EUI 48 address is given by 48 bits.

Addresses are 64 bits. Long addresses tend to increase the transmission time, resulting in increased energy consumption.

Particularly in wireless networks, but also in many wired systems, attention must be paid to energy-saving overall system optimizations. In this regard, it may be advantageous to use shorter addresses that are then unique only in that subarea (link) of a network on the data link layer in which they are used. Another advantage of shorter addresses is that by reducing the transmission time required for this, the medium is occupied for a shorter time and thus, if necessary, the overall data throughput / system performance improves. In most cases addresses which are unique only on a link are assigned centralized by a control unit. Since such address optimizations the global uniqueness of the address in the

As a rule, it does not make sense to specify these shorter addresses during the production process. Rather, the used protocols often provide that the network elements dynamically negotiate the layer 2 addresses to be used only during normal operation.

An example of this is the Bluetooth technology, which is based on the IEEE802.15.1 standard: The "Inquiry procedure" common in Bluetooth offers the possibility of establishing contact between unknown Bluetooth devices by a first device in a network (a "master") so-called "Inquiry" messages sent out. A second device

("Slave"), which could receive an "Inquiry" message based on the frequency used, responds with a so-called "Inquiry Response" message This "Inquiry Response" message contains an identifier that uniquely identifies the second device The identifier can then be used by the first device in a wide The second device can be assigned a unique address on the link, which is therefore suitable for addressing on layer 2.

In principle, the sending of the "Inquiry" message and the "Inquiry Response" message is subject to collision, so if necessary the Inquiry procedure has to be repeated, which means that this process is not deterministic.

With regard to energy supply, the process is not readily transferable to general networks: Bluetooth devices generally have their own self-sufficient energy supply. In contrast, in a communication system in which all subscribers are supplied from a common energy source, the inquiry method can only be used if the accumulated power consumption of all network subscribers is less than the available peak power of the source, taking into account any power losses (for example from a bus ). However, this requirement is not generally met, but can be violated in particular when collisions occur. The probability of violation of the requirement tends to increase with increasing number of participants.

The object of the present invention is to provide ways in which an address can be assigned to network elements (nodes) while minimizing simultaneous access to a communication medium. Disclosure of the invention

The invention proposes a method, a computer program, one or more machine-readable storage media and a communication system having the features of the independent claims 1, 13, 14, 15 respectively.

The inventive method is suitable for an address initialization in a communication system having a control entity and at least one network element. The network element has an identifier that can also be referred to as labeling. For example, it may be a globally unique identifier or may be derived from it, or it may be in the form of a specific latency allocated to the network element; Further possible identifiers are described below.

At a reporting time that depends on the identity of the network element, the network element sends a message to the control entity. According to the method of the invention, an address is also assigned to the network element. Possible forms of the dependence of the reporting time on the identifier are explained below. The reporting time may be fixed at the beginning of the address initialization or may arise during its course, for example depending on queries sent by the control entity; this is explained in more detail below.

The identifier thus determines when the message is sent from the network element to the control entity. This will prevent the network element from sending the message if the control entity sends a message at the same time. If further network elements are contained in the communication system, it can also be avoided (if the functional dependency is appropriately determined) that the network element and one or more of the further network elements transmit messages simultaneously. Thus, the risk of collisions in the address assignment can be reduced.

The method according to the invention may be stored in the form of instructions on one or more machine-readable storage mediums, for example integrated in or used by an embedded system. Alternatively or additionally, it may be implemented in a communication system with a control entity and at least one network element.

Advantageous embodiments are the subject of the dependent claims. The assignment of the address to the network element preferably comprises that in the communication system the address is associated with the network element. This implies that the network element takes over the address and that a link of the network element with the address is stored in the communication system (eg in the control entity).

The control entity preferably has its own address itself. This may be known in advance to the network element (s) in the communication system, for example, or communicated by the control entity, for example together with one or more queries of identifiers or with a start of the address initialization.

If further network elements are present in the communication system, their identifier is preferably of the same type as that of said network element. Examples of possible types are shown below and below:

In one embodiment, the identifier of the network element is a globally unique identification number, such as an EUI-48 or EUI-64

Address. This has the advantage of a simple implementation of the method, because the addresses mentioned are often already established during the production of a network element and can be retrieved from this. In another embodiment, the identifier is a random number representing the

Network element can preferably generate self-sufficient. This embodiment is therefore independent of the presence of globally unique addresses. The probability that several network elements have the same identifier, can be kept small by the random number space is chosen appropriately large depending on the number of network elements.

Further possible forms of the identifier are described below. The control entity can query various possible identifiers by sending query messages to the network element (s) in the communication system. For this purpose, the control entity advantageously has a plurality of possible identifiers. This plurality may be known to the controller or communicated by a higher-level entity, for example, a central entity that manages an overall network in which the communications system is embedded as a subnetwork.

In an advantageous embodiment, the amount is suitably selected, so that the number of identifiers to be queried can be reduced. For example, a 48-bit wide EUI-48 address can be split into a 24-bit manufacturer ID and a 24-bit wide range, which may be distributed individually by the manufacturer. If one can now assume that there are only network elements with a specific manufacturer ID in a network, then it is not necessary to query 2 ⁴⁸ identifiers but only 2 ²⁴ identifiers.

At certain time intervals (clock pulses), the control entity can query with such query messages each one of the possible identifiers. If the identifier of the at least one network element then coincides with the last one queried, the network element sends the message to the control entity at the time of the message. The reporting time is preferably in a time interval of predetermined length from the sending of the last (that is, the matching) query message by the control entity. The time interval can be predetermined such that it allows the network element to receive the corresponding query message, to recognize that the queried possible identifier corresponds to its own identifier and to generate an appropriate message and to send it to the control entity. The time interval is preferably smaller than a polling clock, namely so small that the message is received by the control entity before it sends a new polling message (with a new possible identifier).

As an alternative to a periodic query by the control entity, the network element may also be set up to notify the control entity in a message if the requested identifier does not correspond to its own. The The control entity may be arranged to distinguish such a message from a message (which is sent when the identifier is appropriate) and, moreover, to query the next identifier only when a previous response has been received in the form of such a message or in the form of a message.

Thus, in these embodiments, the network element waits (after starting the method) to send the message until it has received a query message with the appropriate identifier. Thus, the reporting time depends on the identifier of the network element.

In one embodiment, the control entity is arranged to determine the identity of the network element, for example, by receiving the message in a time interval of predetermined length after sending the corresponding (matching) query message. Alternatively, the message itself may include an indication of the identifier of the network element.

If the communication system comprises further network elements, the assignment of the address can take place immediately, that is to say before the sending of a further query message by the control entity. Alternatively, all identifiers of the plurality of identifiers that are known or communicated to the control entity can first be queried before the address assignment takes place. In a further alternative, the control entity performs alternative block dispatching by address mapping, when the control entity has polled a predetermined number of identifiers, or after finding a predetermined number of network elements that have requested identifiers, and after the Address assignment corresponding block by block other identifiers queried and assigned addresses.

The controller entity may be aware of the number of network elements in the communication system or may be notified by a higher-level entity (such as a central entity such as the one mentioned above). In this case, it is advantageous if the control entity stops sending polling messages as soon as it receives a message from each of the network elements Message has been received, with which the corresponding identifier is confirmed directly or indirectly.

In a further embodiment of the present invention, the identification of the at least one network element is given in the form of a specific waiting time which must elapse after a start time before the network element has to send the message to the control entity. The reporting time thus depends on the specific waiting time and thus on the identifier of the network element. The start time can be set, for example, by the control instance or by a higher-level unit.

The waiting time can be determined, for example, from a globally unique identification identifier of the network element or from a random number generated autonomously by the network element. In the latter case, analogous to the above-mentioned by choosing a suitable large random number space, the

Probability to be kept small that several network elements in the communication system have the same identifier.

Alternatively, the identifier of the at least one network element may result from its position having it in the network. For example, you can

Voltage, current or signal losses resulting from the particular distance or cable length characterize the network element due to its position and interpreted as an identifier or transformed into such.

The network can operate with wired and / or wireless technology and / or include an optical system. For example, if the network has a trunk bus or optical fiber, the location with respect to that trunk bus or fiber may be fixed. For example, in a wireless network, the location may define a distance to a base station. In particular, a received signal strength of a pilot signal can be typical for this distance and thus for the network element.

For example, the at least one network element in the ready state can have a certain size (for example voltage, current, signal strength of a pi lotsignals). The position-dependent measurement result can be stored and used as an identifier. In this case, a previous quantization may be useful (in addition to a general A / D conversion) in order to reduce the number of possible values appropriately. This identifier can, as described above, again be polled or interpreted as a random number.

In a preferred embodiment, the reporting time depends on the identifier by being equal to or resulting from a time when a particular minimum voltage or pilot signal strength is applied to the network element, or by decreasing to an interval of a predetermined length falls at that time.

For example, the at least one network element in the ready state can measure a certain size (for example voltage, current, signal strength pilot signal). The size can be changed successively (eg by the central control element). The reporting time can be defined as the point in time at which the measured variable exceeds or falls below a predefined threshold value. The threshold value can be derived in a suitable manner, for example, from a random number or a globally unique identification number.

In the event that the centralized power supply can be regulated by the controller, alternatively the power supply of the communication system can be put under the control of the controller during the address initialization. This can then gradually provide more energy during initialization. If the at least one network element as well as any further network elements present require a certain minimum voltage on the input side (for example due to a particular voltage regulator), a network element can be made operational by a stepwise increase of the input voltage

Report message to the controller and get an address assigned.

Regardless of the type of the respective identifier, the message of the network element sent at the time of the message can indicate the identity of the network element. werkelements included. In this way, the control entity can be informed that the network element is present in the communication system and that the address is being assigned to it. Alternatively, the control entity may be arranged to close the identifier from the time or from the time elapsed from a start time at which it receives the message.

Alternatively or additionally, the message may contain an acknowledgment that the network element is taking over the address. Alternatively, such confirmation may be sent separately in another message.

The message may include an indication of which address the network element has accepted. As a result, higher reliability can be achieved, for example, by simple plausibility checks.

The first address may be assigned to the network element in response to the message by a message from the controller entity. The network element can then confirm the transfer of the address in another message.

Alternatively, the control entity may notify the address to the network element prior to the reporting time. If the communication system comprises one or more further network elements, the control entity can announce the new network element (s) a new address upon receipt of the message. A network element whose reporting time has come can then be assigned this new address.

In one embodiment, the addresses to be allocated are announced in a specific order already at the beginning of the address initialization in the communication system. This is explained in more detail below.

In a preferred embodiment of the present invention, the communication system comprises a plurality of network elements. The above-mentioned (at least one) network element is then a first network element with a first identifier that sends a first message to the controller instance at a first reporting time. The first network element is assigned an address, which is referred to here as the first address. A second network element also has an identifier which, like the first identifier, can have one of the characteristics described above and is referred to here as a second identifier.

The second network element sends according to this embodiment, at a reporting time, which is referred to here as the second reporting time, a message, which is here called a second message, to the control entity. The second network element is assigned an address, which is referred to here as a second address.

The second reporting time depends on the second identifier as described above for the first network element. It can be before or after the first reporting time. Which address is assigned to the first network element and / or which address is assigned to the second network element preferably depends in this embodiment on the time sequence of the two reporting times.

For example, in the communication system (e.g., in the control entity), a list of possible addresses may be stored which are processed in association in a particular order. If the first reporting time is then before the second, the first network element is assigned an address which comes before the second address in the order; The same applies to the opposite case.

The order can be known in advance to the first and the second network element or can be announced, for example, at the beginning of the address initialization. In both cases, if the first reporting time is before the second, the first network element can take over the first address that has not yet been assigned to the order, without the allocation first having to be communicated to it. Preferably, the first network element then sends the first message (or another message, simultaneously or shortly thereafter, that is, before a further step in the Adressinitialisierung is made is sent) also to the second network element, advantageously even to all network elements in the communication system, or to all network elements to which no address has been assigned. In this way, the other network elements know that the named address has already been allocated and that the corresponding next address is assigned to them if their reporting time is the next one.

The assignment of the first address to the first network element can take place immediately after the first message has been sent or can be completed. If the first reporting time before the second, so the assignment can be made in particular before the second reporting time, possibly even before sending another query message by the controller.

Alternatively, the address initialization can be carried out in blocks analogously to the case described above. In particular, in the case that the first reporting time is before the second, the first address can be assigned to the first network element only after the second reporting time.

All of the above embodiments are suitable for use in communication systems having a plurality of network elements, each of which may have a common power supply. They are particularly suitable for address initialization on the data link layer, ie for the case that the address is a layer 2 address (or that the addresses are layer 2 addresses).

However, other applications are possible.

Further advantages and embodiments of the invention will become apparent from the description and the accompanying drawings.

It is understood that the features mentioned above and those yet to be explained below can be used not only in the particular combination given, but also in other combinations or in isolation, without departing from the scope of the present invention. The invention is illustrated schematically by means of exemplary embodiments in the drawings and will be described in detail below with reference to the drawings.

Brief description of the drawings

Figure 1 shows an example of a communication system to which the present invention can be applied.

FIG. 2 shows a possible address initialization with identification request by the control entity.

FIG. 3 shows an address initialization with block processing and identification query.

FIG. 4 shows an address initialization in which the address is sent with the identifier query.

Figure 5 shows an address initialization with address notification in advance.

FIG. 6 shows a possible address initialization in which the identifier is a specific waiting time.

FIG. 7 shows by way of example a typical arrangement of components in a communication system with a linear bus structure.

FIG. 8 shows a possible address initialization which is based on position-specific features of the network elements

FIG. 9 shows an overall network with two separate communication systems in which the invention can be applied.

Embodiment (s) of the invention The communication system 1 shown in FIG. 1 has a control entity 10, a plurality of network elements 101, 102, 103, 104, 105, 110 and a common communication medium 11, via which the network elements and the control entity can communicate with one another. The network elements 101, 102 and 103 have an identifier 1012, 1022 and 1032, respectively, and are assigned addresses 101 1, 1021 and 1031, respectively, in the communication system. The control entity 10 has a set of 100 possible identifiers in the communication system, which it can query if necessary. Furthermore, the control entity shown has its own address 1001 and a set 107 of possible addresses, which it can optionally assign.

FIGS. 2 to 6 and 8 show how the control entity 10 and the network elements 101 and 102 exchange messages in the course of time t. In the associated communication system further, not shown network elements may be present; In an alternative embodiment, fewer network elements may be present than shown in the figures. The respective matching features are provided with the same reference numerals and are explained in the description of Figure 2. In the sequence shown in FIGS. 2 to 5, the respective identifier of the network elements can be, for example, a globally unique EUL address or a random number generated autonomously by each network element. In the latter case, the random number range (from which a random number is drawn) may optionally be suitably limited, for example on the basis of collisions that have occurred from a previous address allocation process and / or on the basis of a (maximum) number of network elements present in the communication system, if this is known.

According to the embodiment illustrated in FIG. 2, the control entity begins to send query messages 201 and 202 to the network elements, in which it successively polls different identifiers to find out which identifiers are present in the communication system. For this, the control entity has a certain area (ie a certain amount) of identifiers available. Is, as in the case of the query message 201, a If the interrogated identifier is not present in the network, the next identifier is continued after a certain time A t (so-called "timeout") If the identifier is present in the network, as in the case of the interrogation message 202, which is received by the associated network element 102 at the time t i, In the case of an embodiment which is not shown, the timeout is thereby triggered again by this message In the form shown, the timeout is a fixed, uniform clock.

Preferably, this clock is selected to be so large that it can be ruled out that the control entity is still sending a new query message during an address assignment.

Based on the feedback 210, the control entity 10 in the message 211 may assign a unique layer 2 address to the network element. The acceptance of this address is confirmed by the network element in the message 212, wherein the confirmation message can also be dispensed with.

The message 210 is sent by the network element 102 at the time of the message T ₁₀ 2. This is done when in a query message (here 202) for the appropriate identifier (here 102) has been asked. Thus, the reporting time T ₁₀ 2 depends on the identifier of the network element 102.

After the expiration of the next time interval A t, the control entity shown sends in the interval cycle further polling messages 203, 204 and 205 in which it in each case asks for a different identifier; As mentioned above, the control entity may alternatively be set up after the address assignment to the

Network element 102 to start with the query of other identifiers, without waiting for the expiration of a fixed time interval. In the query message 205, the identifier of the network element 101 is polled at the time t ₂ . This therefore responds at the time T ₁₀ i, which is thus the reporting time of the network element 101, with the message 213. In it, the network element 101 confirms the identifier. With message 214, the controller instance assigns an address to the network element. In message 215, the network element acknowledges acceptance of the address. This message 215 is optional. In an alternative embodiment, the controller instance does not have to wait for a timeout to proceed with the method. Once the message 212 is received, the process of address initialization for element 102 is complete and the controller may poll the next identifier. If an acknowledgment (message 212) is dispensed with, the control unit can continue to send message 203 directly after the message 21 1 has been sent, without waiting for a timeout

The embodiment of FIG. 3 differs from that shown in FIG. 2 in that the sending of the query messages and the assignment of the addresses are processed one after the other. Thus, a block check is performed. In a first block, all identifiers present in the network are collected by the query messages 201, 202, 203, 204 and 205 as well as the messages 210 and 213. In a second block, the assignment of the addresses and the confirmation of their transfer takes place with the messages 211, 212, 214 and 215.

The confirmation messages 212 and 215 may possibly be waived again.

In an alternative embodiment, a corresponding block processing takes place with inserted address assignment. Not all identifiers available in the network, but only a predetermined number of such identifiers are queried in the first block. In the second block, addresses are assigned to the identifiers collected in this way. In further blocks then other identifiers can be processed analogously.

FIG. 4 shows an embodiment in which the query messages 401, 402, 403, 404, 405 and 406 not only each pass through whether a particular identifier is present in the network, but by default also contain the corresponding address to be assigned next. The network elements 101 and 102 with the corresponding identifier must only confirm in the messages 410 and 411 that the corresponding address has been accepted. FIG. 5 shows an embodiment in which the address to be assigned next is respectively announced in a message 501 or 502 in the communication system. The query messages 201 to 205 then again only need to ask for the respective identifier. If the identifier is suitable, the network elements 102 and 101 send the messages 212 and 215 respectively to the message times T ₁₀ 2 and T ₁₀ i in which they confirm the acceptance of the respective address.

In another alternative, the addresses are assigned in a specific order, which is known in advance to the network elements. In this

In the case shown in FIG. 5, the messages 501 and 502 can be dispensed with.

In both cases, the messages 212 and 215 may contain an indication of the respective identifier. Alternatively, instead of the messages 212 and 215, the messages 210 and 213 described above could be the messages; they contain no confirmation of the address transfer, but an indication of the identifier.

It may be advantageous for the control entity to know the exact number of network elements in the communication system or a maximum number of network elements in the communication system. In these cases, the controller instance may abort the interrogation of the identifiers after the appropriate number of network elements are found. So you do not need to work through the entire space of the identifiers. For example, in the examples above, after the network element 101, the control entity could abort polling of other identifiers if it knows that the number of network elements is limited to two.

The order in which the individual identifiers are queried can be selected differently for different embodiments. Also, the order in which the addresses are assigned can be suitably selected. The amount of addresses to be assigned can be set, for example, to 0. beginning and end in order to keep the necessary address space as small as possible.

Alternatively or additionally, the addresses may be assigned according to a mechanism which ensures that between every two addresses the minimum

Hamming distance becomes maximum (the Hamming distance between two blocks of fixed length (so-called codewords) is the number of different digits). This can reduce the likelihood of misinterpretation of transmission errors.

As a variation of the above-described embodiments, messages may be suitably combined as appropriate, so that while the same information is exchanged, the number of messages required thereto varies.

FIG. 6 shows a message exchange according to a further embodiment of the present invention. The identifier of the network elements corresponds in each case to a specific waiting time. At the time t _0, the control entity starts the address initialization process with the message 600 addressed to all network elements in the communication system. From this

Start signal waits each network element from its identifier corresponding waiting time. The controller instance keeps monitoring the communication channel and waits for incoming messages. It can be advantageous if not only the control entity observes the channel but also the network elements. For example, a network element may then extend the latency by the time that the channel (for previous address assignments of other network elements) is used. In this way, it is possible to prevent a first network element from transmitting and, at a second network element, the waiting time to expire and also to send it. Otherwise, the time interval in which the waiting time is changed by one unit is preferably chosen to be at least so large that at least the messages 601, 602 and 603 can be exchanged. In the illustrated case, the network element 102 has the waiting time Ä After its expiry, the network element 102 transmits the message 601 to the control entity at the time of reporting T ₁₀ 2, thereby reporting the expiration of its own waiting time. The controller then sends in the message 602 the address assigned to the network element 102, whose acceptance confirms the network element in the message 603. Analogously, after expiry of its waiting time Et ₁₀ i, the network element 101 sends a message 604, to which an address is assigned in the message 605, whose acceptance confirms it in the message 606. The confirmation messages 603 and 606 may be omitted.

Preferably, a maximum waiting time t _{M is} provided, after which the process ends. If the control entity knows the number of network elements present in the communication system, the process can alternatively end when all network elements have been found. In this way, the address initialization process can be shortened.

The respective waiting time can for example be calculated on the basis of EUL addresses or on the basis of a random number. It can be provided that not the entire range of identifiers, but only a part of the same is interpreted as waiting time. For example, a 48-bit wide EUI-48 address can be split into a 24-bit manufacturer ID and a 24-bit wide range, which may be distributed individually by the manufacturer. If one can safely assume that there are only network elements with a certain manufacturer ID in a network, then it is not necessary to distribute 2 ⁴⁸ identifiers on the time axis but only 2 ²⁴ identifiers.

It should be noted that under certain circumstances, additional effort for a synchronization between network element and control entity is necessary. It may also be necessary, for example, to take into account in the waiting time of the network elements if communication takes place before the waiting time has elapsed (for example because of the initialization of a network element with a shorter waiting time). With regard to the address space used, the same applies to the above for the figures 2 to 5 executed.

Analogous to the cases described with reference to FIGS. 3 and 5, the embodiments in which the identifier corresponds to a specific waiting time can be combined with an address announcement in advance or else an announcement of the order of the addresses to be assigned.

Finally, it is possible that the control entity from the past time between the start of the award process and the receipt of a message on the identifier of a network element can close (for example, by the introduction of time slots). Thus, the control entity has the information of which network element (with which identifier) it has received a message without an identifier from the network element must be sent explicitly.

FIG. 7 shows a possible arrangement of the network elements in the communication system 1. The common communication medium 11 is a linear bus whose ends are denoted by 120 and 121. The identifier of the network elements is based on their respective position 1043, 1013, 1 103, 1023, and 1053, respectively.

In the event that the centralized power supply 123 can be controlled by the control entity, the power supply of the network can be placed under the control of the control entity during the assignment of the addresses. This allows the controller instance to initialize the

Network gradually provides more and more power to the network, with the result that the network elements will not start up at the same time.

The same applies to the embodiments described above, in which the identifier of the at least one network element results from its position.

Referring to Figure 8, this mechanism is explained in more detail using a voltage source 123 (as shown in Figure 7), for example: From a start signal 800 at time t ₀ , the control instance starts to increase the voltage (slowly), for example beginning with OV. Since the individual network elements are distributed on the bus and therefore there is a different distance (cable length) between each network element and the power supply 123, different voltage losses occur on the cable, whereby a different voltage is applied to each network element on the input side. Furthermore, it can be assumed that each network element has a voltage regulator, which is only ready for operation when the input side of the network element (and thus the voltage regulator) is applied a certain minimum voltage. Since the losses on the cable cause the input voltage to be different for each network element, the slow increase in the input voltage ensures that the network elements do not simultaneously but successively (in terms of time) have the required input voltage and are therefore ready for operation one after the other are.

In the illustrated case, the network element 101 has the required input voltage at the instant ΤΊ (which thus becomes the reporting time of the network element). The network element 101 therefore sends a message 801 to the control entity 10. It then assigns to the network element 101 in a message 802 an address whose adoption confirms the network element 101 in the message 803.

At a later time T ₂ , the network element 102 has the input voltage required there. Accordingly, this network element also sends a message 804 to the control entity, is assigned an address in the message 805 and confirms its acceptance in the message 806.

The messages 803 and 806 with confirmation of the address assignment can be dispensed ter circumstances.

In order to prevent multiple network elements from becoming operational at the same time, the voltage controlled by the control entity 10 must be increased in correspondingly small steps and / or, if appropriate, a minimum distance (minimum cable length) between two network elements, so that the voltage loss caused by the line has a corresponding effect. As soon as a network element is ready for operation due to the voltage present on the input side, it can send a message to the control instance and thus be assigned an address by it.

The control entity can alternatively or additionally be set up, with an increase of the voltage the switching on of a network element on the basis of an increased power consumption or an increased voltage drop or the like. without, or before, sending a message from the network element to the control entity.

This Adressinitialisierungsprozess can be modified analogously to the methods described above by the identifier of the transmitting network element is sent in the messages 801 and 804 and a block handling takes place in which several or all addresses are assigned to a collection of several or all existing identifiers.

Alternatively or additionally, the input voltage (or the input current) can also be measured (with quantization) at each network element, for example. The measured value can then also be interpreted as identifier or random number, so that the approaches described above can be used.

Particularly in the case of wireless communication, instead of the voltage / current measurement, the received signal strength of a pilot signal can be measured. This measured value can also be interpreted as identifier or random number, so that the above approaches can be applied analogously, if appropriate again with appropriate quantization.

In further embodiments, it is assumed that a central entity (for example the control entity) in the network is aware of which network elements (for example characterized by a network node that is unique in the network) Identifier, eg serial number or EUI-48 / EUI-64 address) are part of the network. The corresponding identifiers of the network elements may possibly be collected during the installation process or possibly even through it.

According to such an embodiment, the control entity is informed which identifiers are located in the network. Once the controller instance has this information, it can directly assign an appropriate address to the network elements. In the structure shown in FIG. 9, an overall network 90 is divided into a plurality of subnets 1000, 2000, each subnet being its own

Control instance 10 or 20 has. The individual control instances are connected to each other via a suitable communication interface (eg a backbone) 92. The controller 10 may include a block 107 with possible addresses as shown in FIG. Alternatively or additionally, the control entity 20 may contain an analog block with possible addresses that it can allocate. Additionally or alternatively, the central element 91 may contain a block with possible addresses and specify to the control authorities 10 and 20 which addresses from this block they can allocate. In either case, the address ranges for networks 1000 and 2000 may include one or more common addresses.

In such a network, there are various ways of telling the control entity 10, 20 which identifiers are available in the network 100 or 200: a) Each control entity is informed directly exactly the identifiers that exist in its subnetwork. b) An optional existing central element 91 is communicated all identifiers from the entire network and there is an assignment, which

Identifier is located in which subnet. The central element tells each control instance exactly the identifiers that also fall into its subnet. c) Each controller instance will be notified directly of all identifiers from the entire network. d) An optionally available central element is notified of all identifiers from the entire network, and the central element forwards to each control entity all identifiers that exist in the overall network.

The role of the central element can also be assumed by a control entity or the central element can also have additional control entity functionality.

While the control entity in variants a) and b) can assign suitable addresses directly to the network elements, variants c) and d) have the peculiarity that in a subnetwork not all identifiers of the overall network are necessarily present, so that the control entity again " must "try what identifiers are actually present in their network." This can in principle be accomplished by the concepts set out in relation to Figures 2-8.

When using the variants, in which query messages are sent, there is a time advantage, since the search space is correspondingly limited and not the entire range of identifiers must be queried.

In addition, in this case there is the advantage that the individual control entities can communicate with each other via the backbone. This allows the controller on the network to tell each other which identifiers have already been found. This will prevent a controller instance from looking for an identifier that has already been found in another subnet. So a temporal advantage can be achieved. This effect can additionally be influenced by giving each individual control entity a different start value (with which the interrogation of the identifiers is to be started) and / or another search strategy (eg search order).

Claims

claims

A method for address initialization in a communication system (1), comprising a control entity (10) and at least one network element (101, 102, 103, 110), the network element being identified by an identifier (1012, 1022, Et ₁₀ i, At ₁₀₂ ),

and wherein the method comprises the following steps:

sending a message (210, 213, 410, 601, 604, 801, 804) from the network element to the control entity,

wherein the sending takes place at a reporting time (T ₁₀ i, T ₁₀ 2), which depends on the identifier of the network element and

assigning an address (101 1, 1021, 1031) to the network element.

2. The method of claim 1, wherein the identifier of the at least one network element is a globally unique identifier or a specific, preferably autonomously generated by the network element random number, and wherein the message is preferably an indication of the identifier and / or the address and / or that the address is accepted by the network element.

The method of claim 1 or 2, further comprising sending at least one query message (201, 202, 203, 204, 205) by the control entity, wherein the query message retrieves a possible identifier of the network element and preferably further includes an indication of the first address ,

4. The method according to claim 3, wherein the possible identifier is selected from a predefined set (100) of possible identifiers which are requested successively,

and / or wherein the control entity sends an indication (501, 502) of the first address to the network element before sending the at least one query message.

5. The method according to claim 3, wherein the potential identifier requested in the query message matches the identifier of the network element,

and wherein the reporting time depends on the identifier of the network element by falling within a time interval of predetermined length from the time (ti, t ₂ ) at which the network element receives the query message.

6. The method of claim 1, wherein the identifier of the at least one network element a specific waiting time (Ät ₁₀ i, Ät ₁₀ 2) is calculated, for example, a globally unique identifier or a specific, preferably independently generated by the network element random number of the network element or became;

and preferably wherein the reporting time (Τ ₁₀ ι, T102) depends on the identifier of the network element by being an expiration time of the specific waiting time after a start time (t ₀ ) or falls within a time interval of predetermined length from the expiration time,

wherein the start time was set, for example, by the control entity, and / or wherein the control entity preferably before the start time sends an indication of the address to the network element.

The method of claim 6, wherein the message includes an indication of

that the specific waiting time has expired

and / or which globally unique identifier (1012, 1022) the

Network element owns

and / or that the address is accepted by the first network element and / or is the address.

8. The method according to claim 1, wherein the identifier of the at least one

Network element, a specific position (1013, 1023, 1043) of the network element in the network characterized.

9. The method of claim 8, wherein the reporting time depends on the identifier of the network element by falling within a time interval of predetermined length from the time point (ΤΊ) to which a predetermined minimum voltage or current or received signal strength is applied to the network element.

and preferably wherein the message includes an indication of

that the time has been reached or exceeded

and / or that the predetermined minimum voltage or current or received signal strength is present

and / or that the address is accepted by the network element

and / or what the address is.

10. The method according to claim 1, wherein the network element is a first network element, the identifier is a first identifier, the message is a first message, the reporting time is a first message time, and the address is a first address.

wherein the communication system further comprises at least a second network element having a second identifier,

the method further comprising:

sending a second message from the second network element to the control entity and at a second reporting time and

assigning a second address to the second network element;

wherein the assignment of the first address and / or the assignment of the second address is based on a chronological order of the first and second reporting times,

wherein the first and second network elements are preferably powered by a common, centralized power source (123).

1. The method according to claim 10, wherein sending the second message (213, 215, 604, 804) after sending the first message (210, 212,

601, 801), and wherein the first address is assigned to the first network element before the second message is sent, wherein preferably the control entity sends an indication of the second address to the second network element before sending the second message (502)

or wherein the first address is assigned to the first network element after the second message has been sent, wherein the method is preferably according to claim 4 and the assignment of the first address after querying all possible identifiers of the predetermined amount (100) or after querying a predetermined number Identifiers from the specified quantity (100) takes place.

12. The method according to any one of claims 10 or 1 1, wherein the first and the second address and the temporal order on which the assignment of the addresses based, the first and / or the second network element before sending the first and / or second message are known

and wherein the first message includes an indication of which address the first network element takes,

and preferably wherein the first message is also sent to the second network element.

13. A computer program with program code means which, when executed on a computer or a corresponding arithmetic unit, cause the computer or the corresponding arithmetic unit to perform those steps in accordance with a method according to one of claims 1 to 12, which is performed by a control entity in one Are to execute communication system; and / or to carry out those steps according to a method according to one of claims 1 to 11, which are to be executed by a network element (101, 102, 103, 110) in a network.

14. A machine-readable storage medium with a computer program stored thereon according to claim 13.

15. A communication system (1) having at least one network element (101, 102, 110), a control entity (10) and a common communication medium (11), wherein the communication system is adapted to carry out one of the methods of claims 1 to 12.