DE102019213322A1 - Ethernet physical layer transceiver for two-wire bus topology - Google Patents

Abstract

A physical layer transceiver (100) set up to transmit and receive Ethernet data frames via a two-wire bus (102) shared by several network devices comprises a driver circuit (104) which is set up to connect each of the two lines of the two-wire bus (102 ) actively to transfer data to a first of at least two electrical states, the at least two electrical states representing signals by means of which the data are encoded. The physical layer transceiver also includes a receiver circuit (106) which is set up to sense electrical states present on the two lines of the two-wire bus (102) and to derive signals therefrom which encode received data. Furthermore, the physical layer transceiver comprises a logic circuit (108) which is set up to carry out an arbitration of an active bus access based on the priority of a message to be sent. The logic circuit (108) is set up to control the driver circuit (104) and the receiver circuit (106) to carry out the arbitration at a first data rate which allows the resolution of collisions between signals from two network devices that are transmitting at the same time. The logic circuit (108) is also set up to control the driver circuit (104) to transmit an Ethernet data frame to be sent at the first data rate or a second data rate which is greater than the first data rate if the result of the arbitration permits transmission.

Description

FIELD

The present invention relates to a physical layer transceiver for digital networks, which transmits data in Ethernet packets according to the IEEE standard 802.3, in particular to a physical layer transceiver for networks in which several network devices are directly connected to the multi-master principle a shared data bus are connected. The present invention also relates to a network device with a physical layer transceiver according to the invention, a control method of the network device for the transmission of data, and a vehicle with one or more such network devices.

BACKGROUND

Digitally networked devices are used today in more and more application areas, for example to be able to offer known functions with less effort than before or to provide new functions that are only possible through the interaction of several devices. The networking often takes place according to different standards using different transmission protocols and hardware.

In land, water and air vehicles, among other things, communicatively networked digital control units have been used for a long time, which has led, among other things, to a significant reduction in the number of lines previously used to directly connect individual devices to one another, which adds weight and costs could be saved. In a large number of today's systems, digital communication interfaces are used according to the most varied of standards, which over time have proven to be sufficient and reliable for certain control units for the respective purpose. Networking of components according to the CAN bus standard, which defines the physical layer and the data security layer of the OSI layer model, is inexpensive and widespread in the automotive sector. Networking according to the Local Interconnect Network Standard (LIN or LIN bus), which is less versatile than the CAN bus and which only allows a lower data rate, is also inexpensive and widespread in the automotive sector. Networking according to the FlexRay standard is used in particular when a higher data transmission rate, real-time capability and reliability are required, e.g. for drive-by-wire systems. However, the focus is often on the higher data rate, which is necessary today due to the continuous increase in driver assistance systems in the drive and chassis area in premium vehicles. Like the CAN bus and LIN bus standards, FlexRay defines the physical layer and the data security layer in the OSI layer model.

Progress in the miniaturization of components for digital communication interfaces and ever increasing computing capacities even with simple microprocessors, the associated desire to provide more and more data from sensors for monitoring and control for each system, as well as the requirements for more and more additional functions and Services can cause the networks or components provided in some systems to reach the limits of their performance. If a component of a networked system was previously networked via a comparatively slow network such as the LIN-BUS, for example, a complex adaptation must be made so that the use of a faster network is possible. This adaptation itself can be laborious and expensive, and the faster networking can require the use of more expensive and manufactured in smaller numbers components. In addition, it may be necessary to change the cabling, for example in the case of a vehicle network, so that further expenditure arises here.

Primarily for reasons of cost, it is preferred to use identical or at least based on identical hardware digital control and monitoring units in a large number of different systems. Especially with devices that are networked with one another by means of digital networks, and especially with devices that are subject to time-consuming and expensive approval processes, it can make sense to consider the possibility of simple adaptation to future systems at the design stage.

In the area of tension between the requirement for particularly powerful communication interfaces on the one hand and particularly cost-effective communication interfaces for simple systems on the other hand, the use of standardized components and, in particular, standardized software for their operation is one way of meeting both requirements with reasonable economic effort. At the same time, the use of as many standardized components as possible can keep the effort for adjustments small.

For example, interfaces according to the IEEE 802.3 standard have become established in many areas for digital communication between devices, which are also known under the name Ethernet, although there is no complete correspondence between that in IEEE 802.3 Standard and the specifications often summarized under the name Ethernet. In the following, the name Ethernet and the IEEE Standard 802.3 are used synonymously.

The collective term Ethernet specifies software, protocols, etc., and hardware for wired data networks, which were originally intended for local data networks. Ethernet enables data exchange in the form of data frames between the devices connected in a local area network (LAN).

Original versions of the Ethernet used coaxial cables for the transmission of the data, later versions used two or more pairs of twisted pairs. Both coaxial cables and cables with twisted pairs required the use of complex cables, plugs and sockets, which are expensive and unsuitable for harsh environmental conditions, such as those found in all types of vehicles in particular. The development of plugs and sockets adapted to this, in turn, negates the cost advantage resulting from the use of standardized components. In addition, the disadvantage of the special cables remains, so that, despite the many advantages of the availability of standardized software and hardware components, Ethernet connections are often only considered for connections that require a high data rate. In addition, these connections are usually designed as point-to-point connections, which require active components for switching to connect more than two devices. These active components themselves require energy for their operation and represent weak points with regard to the reliability of the entire network or larger parts of it.

Proceeding from this, the present invention has the object of creating a possibility of connecting system components for the transmission of Ethernet data frames by means of a two-wire bus in order to overcome or at least improve one or more of the above-mentioned problems.

DESCRIPTION

To solve the problem, the invention proposes, according to a first aspect, a physical layer transceiver for transmitting and receiving Ethernet data frames via a two-wire bus shared by several network devices, which is set up to allow arbitration based on the priority of a message to be sent an active bus access.

According to a second aspect, the invention proposes a network device which has a physical layer transceiver according to the first aspect and is thus set up for the transmission of Ethernet data frames via a two-wire bus shared by several network devices.

According to a third aspect, the invention proposes a control method of a network device for the transmission of Ethernet data frames over a two-wire bus shared by several network devices, in particular for priority-oriented arbitration for the transmission access to the bus.

An Ethernet data frame according to the IEEE Standard 802.3 begins with a 6 byte long MAC address of the recipient of a message, followed by the 6 byte long MAC address of the sender. An Ethernet packet contains an Ethernet data frame preceded by a 7-byte preamble and a 1-byte start frame delimiter (SFD). The use of the preamble enables bit synchronization of the devices connected to a network. This is particularly advantageous when there are no permanently synchronized point-to-point connections. The terms Ethernet data frame and Ethernet packet are used synonymously in this description, provided that the respective context does not result in any other use.

A physical layer transceiver according to the first aspect, which is set up to transmit and receive Ethernet data frames via a two-wire bus shared by several network devices, comprises a driver circuit which is set up to actively transfer each of the two lines of the two-wire bus into one to set the first of at least two electrical states, wherein the at least two electrical states represent signals by means of which the data are encoded during transmission. Electrical states can be electrical potentials, for example. The two lines of the two-wire bus are set up to strive for a second of the at least two electrical states if no network device connected to the bus actively puts one of the lines of the bus into another state. This can be done, for example, by resistors connected to the lines, which connect the lines to an electrical potential.

The physical layer transceiver according to the first aspect further comprises a receiver circuit which is set up to sample electrical states present on the two lines of the two-wire bus and to derive signals therefrom which encode received data.

The physical layer transceiver according to the first aspect also comprises a logic circuit which is set up to carry out an arbitration of an active bus access based on the priority of a message to be sent. The logic circuit is set up to control the driver circuit and the receiver circuit for carrying out the arbitration at a first data rate which allows the resolution of collisions between signals from two network devices that are transmitting at the same time. The logic circuit is also set up to control the driver circuit for the transmission of an Ethernet data frame to be sent at the first data rate or at a second data rate which is greater than the first data rate if the result of the arbitration permits transmission. A transmission is normally permitted if the priority of a separate message to be sent is higher than the priorities of the messages to be sent by other network devices connected to the shared bus. The logic circuit can, for example, comprise a finite state machine, or a microprocessor that executes computer program instructions that implement the function of the logic circuit.

In one or more refinements, the physical layer transceiver is set up to carry out bit-by-bit arbitration in which each sending physical layer transceiver monitors the bus while it is itself sending an identifier containing the priority of its message to be sent. If two or more physical layer transceivers send at the same time, the first dominant bit of one of the physical layer transceivers overwrites the corresponding recessive bit of the other physical layer transceiver (s), which they recognize and then terminate their attempt to transmit. A physical layer transceiver can switch to a passive bus monitoring mode as soon as it can be seen that it is not allowed to transmit after the arbitration has ended. In the present case, a recessive bit is a bit that is represented by an electrical state in which a line is not actively switched to a different electrical state by the driver circuit of one of the physical layer transceivers connected to the bus. The state belonging to a recessive bit can be set, for example, by connecting a line to an electrical potential via a resistor. A state belonging to a dominant bit is then achieved by actively setting a potential of the line by the driver circuit. In most cases, a physical layer transceiver will be left at the end, whose message to be sent has a higher priority than that of the other physical layer transceivers. Should two or more physical layer transceivers ever have messages to be sent with the same priority, other methods for collision resolution can also be used.

In one embodiment, the physical layer transceiver is set up to receive a priority of a message to be sent, which is required for arbitration, via a control interface. The control interface can, for example, provide communication with a MAC or another component of the network device that is arranged on a higher layer of the OSI layer model. An example of such a control interface is a Serial Peripheral Interface (SPI), which is present in many integrated digital circuits for configuration or other data exchange with other circuits. Alternatively, the physical layer transceiver can have a fixed priority that applies to all messages to be sent.

In one embodiment, the physical layer transceiver is set up to derive a priority of a message to be sent, which is required for arbitration, in a header in accordance with the IEEE standard 802.1Q of the Ethernet data frame. For this purpose, the physical layer transceiver can have a buffer memory which is sufficiently dimensioned to store at least one Ethernet data frame to be sent. The physical layer transceiver can then analyze the buffered Ethernet data frame and read out the priority information contained in a header in accordance with IEEE Standard 802.1Q and use it for the arbitration. If the Ethernet data frame does not have such a header, which can be recognized by the lack of the EtherType value 0x8100 reserved for this header in the 21st and 22nd bytes of the Ethernet data frame, the physical layer transceiver can either set a lowest priority for the accept sending message, or a priority set for the physical layer transceiver.

In one or more refinements, the physical layer transceiver is set up to encode symbols on the two-wire bus with more than two states on one or both lines. In this case, the driver circuit can be set up to encode symbols to be sent during the arbitration with two states on the lines and, after the arbitration, to encode symbols with more than two states in order to send the message.

In one or more refinements, the physical layer transceiver is set up to use the scanned states on the lines of the two-wire bus to recognize a current occupancy of the bus and to postpone arbitration until no signals are transmitted on the bus. A physical layer Transceivers, for example, monitor all communication on the bus and compare the block check string sent at the end of an Ethernet data frame with the self-calculated checksum of the data frame. The interpacket gap following the block check character sequence, which designates the minimum time interval between two packets sent on the transmission medium, can then be used to start an arbitration. The minimum waiting time between two sent Ethernet packets is generally 96 bit times, at 10 Mbit / s this corresponds to 9.6 µs. Alternatively, arbitration can be postponed until an arbitration message sent by another network device is detected.

In one or more embodiments, the physical layer transceiver is set up to signal a result of an arbitration and / or a bus monitoring to the components of the data link layer, in the case of an occupancy of the bus after the arbitration by another network device, a message to be sent by a component belonging to a higher layer of the OSI layer model than the physical layer transceiver is stored or is discarded until a new transmission attempt is made.

A network device set up for the transmission of Ethernet data packets via a two-wire bus shared by several network devices according to the second aspect has a microprocessor and associated memory. The network device is also set up to send and receive Ethernet data frames. For this purpose, the network device has a Logical Link Controller (LLC) and a Media Access Controller (MAC), which implement the data link layer of the OSI reference model. In addition, the network device has a physical layer transceiver according to the invention which implements the physical level of the OSI reference model and which, as described above, is set up for arbitration, for sending and receiving signals on a two-wire bus. The physical layer transceiver provides an adapted interface for the two-wire bus. The physical layer transceiver and the media access controller are communicatively connected to one another via a media-independent interface (MII).

A control method of a network device for the transmission of Ethernet data frames over a two-wire bus shared by several network devices includes the sending of a first arbitration message at a first data rate, which allows the resolution of collisions between two network devices that are sending at the same time, the first arbitration message providing information about Contains priority of a message to be sent by the network device. While the first arbitration message is being sent, it is also monitored whether a second network device connected to the shared two-wire bus is sending a corresponding second arbitration message at the same time. If the priority of the message to be sent by the second network device is lower than that of its own message to be sent, the message is sent at the first data rate or at a second data rate which is higher than the first data rate. If the priority of the message to be sent by the second network device is higher than that of its own message to be sent, the network device is triggered to monitor the two-wire bus for the end of a transmission of a message sent with a higher priority.

In one embodiment, the arbitration message is only sent when either no other network device connected to the shared two-wire bus is sending, or when another network device sends an arbitration message via the jointly used two-wire bus.

Preferably, only one Ethernet data packet is sent from each transmitter, so that a new arbitration can take place after it has been transmitted. It is of course also possible for an Ethernet data packet to be sent without prior arbitration, e.g. if messages are only sent sporadically over the shared two-wire bus, and especially if they are sent at comparatively regular intervals, so that times , for which no messages are normally sent over the bus, can easily be determined or estimated. A corresponding evaluation of the network traffic can take place in each network device, and corresponding information can be passed on to the network interface or the physical layer transceiver with each message to be sent, so that the arbitration can be omitted in individual cases.

The arbitration based on the priority of the messages to be sent creates a hierarchy of messages among one another. For the transmission of time-critical messages, an identifier of high priority can be assigned in order to give them priority during the transmission. As a result, messages can be transmitted while taking into account a desired quality of service in accordance with IEEE802.3Q. However, even with high-priority messages, the time of transmission cannot be determined precisely in advance, because messages currently being transmitted cannot be interrupted and the Delay the start time of a broadcast up to the maximum message length. Only the maximum transmission delay for the message with the highest priority can be determined if the maximum message length is known. For messages with low priority, it is generally not possible to make a statement about the time of sending.

The use of a simple two-wire bus enables the use of comparatively simple plugs and sockets, which have also been used in vehicles of all kinds for a long time and have proven their reliability under many environmental conditions. To reduce susceptibility to interference and radiation, the two conductors of the two-wire bus can be twisted. Since the two-wire bus enables the cost-effective connection of devices, the use of the Ethernet protocol is also of interest for the networking of devices that may currently still transmit little data, for which other network types would be sufficient, but for which it is expected that they are will send larger amounts of data in the future.

The separation of the functions of the Logical Link Controller, Media Access Controller and Physical Layer Transceiver as well as the connection of these function modules by means of a media-independent interface makes it possible to keep the changes to the control unit small when using a control unit in a different network structure. For example, simply by exchanging the physical layer transceiver, it can be adapted to use in a high-speed Ethernet network with one or more switches that ultimately provide point-to-point connections. An adaptation in higher layers of the OSI layer model is not necessary because these already provide correct Ethernet data frames. Adaptation to a higher data rate is also easily possible in this way if the maximum data rate that can be achieved by means of the twisted wire pair of a jointly used two-wire bus no longer meets the new requirements. Two-wire buses with twisted cables can be used up to a data rate of around 10 Mbit / s. If higher data rates are required in a different network configuration, this can be easily implemented for a network device by exchanging the physical layer transceiver.

The existing physical layer transceiver according to the first aspect of the present invention can also be used for point-to-point connections without adaptations if the achievable data rate is basically still sufficient, but the increasing competition from an ever larger number of network devices connected to the same bus leads to an increase in access collisions, so that active network components are used to switch the data streams between network devices. The time required for the arbitration is usually negligible compared to the time required for the transmission of the Ethernet data frames, so that this does not result in any significant disadvantage in the case of a fixed point-to-point connection. Alternatively, it is of course possible for the physical layer transceiver to recognize a fixed point-to-point connection and then omit the arbitration, or for the physical layer transceiver to receive a corresponding signal in response to which the arbitration does not apply.

Although the invention has been described in the present description primarily with reference to an application in vehicles, that is to say cars, trucks, aircraft or rail vehicles, an application in other areas is also conceivable, for example in industrial automation.

Figure list

• 1 eine vereinfachte schematische Darstellung eines exemplarischen erfindungsgemäßen Physical Layer Transceivers,
• 2 ein Netzwerk mit mehreren an einen Zweidraht-Bus angeschlossenen Netzwerkgeräten, die jeweils eine erfindungsgemäße Physical Layer Transceiver aufweisen,
• 3 ein vereinfachtes Flussdiagramm eines erfindungsgemäßen Ansteuerverfahrens eines Netzwerkgerätes zur Übertragung von Ethernet-Datenframes über einen Zweidraht-Bus, und
• 4 eine schematische Darstellung von Signalen auf einer Busleitung während der Arbitrierung.

The invention is described below with reference to the drawing. In the drawing shows

• 1 a simplified schematic representation of an exemplary physical layer transceiver according to the invention,
• 2 a network with several network devices connected to a two-wire bus, each of which has a physical layer transceiver according to the invention,
• 3rd a simplified flow diagram of a control method according to the invention of a network device for the transmission of Ethernet data frames via a two-wire bus, and
• 4th a schematic representation of signals on a bus line during arbitration.

Identical or similar elements can be referenced in the figures with the same reference symbols.

DESCRIPTION OF EXEMPLARY EMBODIMENTS

1 shows a simplified schematic representation of an exemplary physical layer transceiver according to the invention 100 . A driver circuit 104 and a receiver circuit 106 of the physical layer transceiver 100 are with a two-wire bus 102 connected. There is driver circuit 104 set up for this purpose, the lines of the two-wire bus 102 actively put into a first of at least two electrical states. Which represent at least two electrical states Signals used to encode the data. The receiver circuit 106 is set up for this on the two lines of the two-wire bus 102 to sense electrical conditions present and to derive signals therefrom which encode received data. Physical Layer Transceiver 100 also has a logic circuit 108 on which with driver circuit 104 and receiver circuit 106 is communicatively connected and controls it. Logic circuit 108 controls driver circuit 104 and receiver circuit 106 among other things, to carry out an arbitration of an active bus access based on the priority of a message to be sent. The information relating to the priority of a message to be sent can be sent to the logic circuit 108 e.g. via an interface 110 be supplied, or from the via an interface 114 received data of a message to be sent are extracted. To do this, the message can be stored in a buffer memory 112 are temporarily stored in order to read out the priority information. The alternatives are indicated in the figure by dashed lines. interface 114 can use physical layer transceivers 100 connect with elements of a higher layer of the OSI layer model, for example a media access controller. The data format sent via this media-independent interface is preferably independent of a data format for transmission via the medium used.

2 shows a network with several network devices connected to a two-wire bus 200 , 201 , 202 each having a physical layer transceiver according to the invention 100 exhibit. Physical Layer Transceiver 100 each of the network devices 200 , 201 and 202 is with a media access controller 204 connected, which in turn is connected to a Logical Link Controller 206 connected is. Media Access Controller 204 and Logical Link Controller 206 implement elements of the data link layer of the OSI layer model. Logical Link Controller 206 is connected to elements of higher layers of the OSI layer model via interface 208 for the reception of data to be sent and for the transmission of received data (not shown in the figure). The lines of the two-wire bus 102 are terminated at the ends. The termination 302 can be formed by individual resistors that are connected between the lines at each end of the two-wire bus, but termination with two resistors of the same size, connected in series between the two lines, between which a capacitor to ground potential or Ground is switched. This circuit, also known as a split bus termination, acts like a low-pass filter. Without influencing the DC voltage ratios, high-frequency signals are short-circuited to ground via the capacitor, which improves interference immunity and also reduces emissions.

3rd shows a simplified flow diagram of a control method according to the invention 400 a network device for the transmission of Ethernet data frames via a two-wire bus. In step 402 the network device sends a first arbitration message with a first data rate, which allows the resolution of collisions between two network devices that are transmitting at the same time. The first arbitration message contains information about a priority of a message to be sent by the network device. Sending takes place by applying appropriate signals to the lines of the two-wire bus. During the transmission, the lines of the two-wire bus are monitored for this, step 404 whether a second network device connected to the shared two-wire bus is sending a corresponding second arbitration message at the same time and whether the priority of its message to be sent is higher than that of its own message to be sent. If this is the case, "yes" branch from step 405 , the two-wire bus is monitored, step 408 to identify the end of a transmission of a message sent with a higher priority. If the priority of your own message to be sent is higher than that of the messages to be sent from all other network devices, "no" branch of step 405 , your own message will be in step 406 at the first data rate or at a second data rate which is higher than that sent as the first data rate. The arbitration can be preceded by a monitoring of the two-wire bus, step 401 , by means of which it is determined whether another network device is currently actively accessing the bus, i.e. transmitting, or whether another network device is initiating an arbitration.

4th shows a schematic representation of signals on a bus line during arbitration. In the top line of the figure, data blocks are shown, the bits of which are mapped onto the bus line one after the other in the form of corresponding electrical states. In order to make the chronological sequence easier to understand, the bits or states on the bus are shown one after the other, starting from the left.
The second to fourth lines are those of network devices 200 , 201 and 202 signals applied to the line. Logical “Iow” levels are dominant states, ie the driver circuit of the network device shorts the line to a lower reference potential. Logical "high" levels are automatically set on the line by appropriate pull-up resistors if no driver circuit of a network device actively short-circuits the line to the lower reference potential. The pull-up resistors can be provided centrally for the bus line, or in each of the driver circuits; The only thing that matters is that the logical “low” level can always be set reliably within the required time span.
In the lower line, the levels on the bus line are shown, which result from the simultaneous activation of the bus line BUS through the three network devices 200 , 201 and 202 surrender.
Starting from a logical “high” level in the idle state, each of the three network devices 200 , 201 and 202 signals that you want to transmit a message over the bus. To do this, the physical layer transceivers of the network devices short the line to the lower reference potential to generate the start arbitration signal SAS to send. Then each of the network devices begins to communicate the priority of its message over the bus. In the example, 10 bits are provided for this. The bits with the numbers 10 - 6 are on all three network devices 200 , 201 and 202 identical, so that none of the network devices can determine during bus monitoring that other network devices are also actively changing the electrical status of the line at the same time. In bit number 5, the network device “sends” 201 a logical “high” level while the network devices 200 and 202 Actively bring the line to a logical "low" level. Network device 201 recognizes this as a higher priority of another network device and then switches to network monitoring, i.e. no longer modulates the electrical states on the line itself in order to detect the end of a pending transmission. This is indicated in the figure by the gray background of the line. Network devices 200 and 202 continue to send their respective message priority because so far neither of the two network devices has been able to detect that another network device is also sending. The network device only “sends” bit number 2 200 a logical “high” level while network device 202 actively brings the line to a logical "low" level. Now also recognizes network device 200 that another network device wants to send a message with a higher priority and switches the network device as before 201 to a network monitoring in order to recognize the end of a pending transmission, again indicated in the figure by the gray background. Network device 202 is the only one left with the highest priority and sends address information and data after a signal EAS , which indicates the end of the arbitration phase and the beginning of the actual Ethernet data frame.

List of reference symbols

100

Physical Layer Transceiver

102

Two-wire bus

104

Driver circuit

106

Receiver circuit

108

Logic circuit

110

Interface for receiving priority information

112

Buffer storage

114

media-independent interface

200

Network device

201

Network device

202

Network device

204

Media Access Controller

206

Logical Link Controller

300

network

302

Bus termination

400

Procedure

402-408

Procedural steps

SAS

Arbitration start signal

EAS

Arbitration end signal

BUS

Bus line

Claims (13)

Physical layer transceiver (100) set up to transmit and receive Ethernet data frames via a two-wire bus (102) shared by several network devices, comprising: - A driver circuit (104) which is set up to actively put each of the two lines of the two-wire bus (102) into a first of at least two electrical states for the transmission of data, the at least two electrical states representing signals by means of which the Data is encoded, - A receiver circuit (106) which is set up to sense electrical states present on the two lines of the two-wire bus (102) and to derive signals therefrom which encode received data, - A logic circuit (108) which is set up to carry out an arbitration of an active bus access based on the priority of a message to be sent, the logic circuit (108) being set up to carry out the driver circuit (104) and the receiver circuit (106) to control the arbitration at a first data rate, which allows the resolution of collisions of signals from two network devices that are transmitting at the same time, and wherein the logic circuit (108) is also set up to control the driver circuit (104) for the transmission of an Ethernet data frame to be transmitted at the first data rate or to control a second data rate, which is greater than the first data rate, if the result of the arbitration allows a transmission. Physical Layer Transceiver (100) according to Claim 1 , also set up to receive a priority of a message to be sent, which is required for arbitration, via a control interface (110). Physical Layer Transceiver (100) according to Claim 1 , wherein a buffer memory (112) is provided which is set up to buffer at least one Ethernet data frame to be sent, and wherein the physical layer transceiver (100) is further set up to set a priority of a message to be sent required for arbitration in a header according to the IEEE standard 802.1 Q of the Ethernet data frame. Physical layer transceiver (100) according to one of the preceding claims, wherein the driver circuit (104) is set up to encode symbols of a message with more than two states on one or both lines of the two-wire bus (102). Physical Layer Transceiver (100) according to Claim 4 , wherein the driver circuit (104) is set up to encode symbols to be transmitted during the arbitration with two states on the lines, and after the arbitration to encode symbols with more than two states in order to send the message. Physical layer transceiver (100) according to one of the preceding claims, also set up to recognize a current occupancy of the bus based on the scanned states on the lines of the two-wire bus (102) and to postpone arbitration until no signals are transmitted on the bus . Network device (200) set up for the transmission of Ethernet data frames via a two-wire bus to which two or more network devices are directly connected for shared use, comprising: a logical link controller (202) and a media access controller (204), which implement the data link layer of the OSI reference model, - a physical layer transceiver (100) according to one of the Claims 1 - 6th , which implements the physical layer of the OSI reference model, wherein the media access controller (204) and the physical layer transceiver are communicatively connected to one another via a media-independent interface (206), and wherein the physical layer transceiver (100) is an adapted interface for the Two-wire bus (102) provides. Network (300) with two or more network devices connected via a two-wire bus Claim 7 wherein the two-wire bus is an end-terminated (302) twisted pair bus. Control method (400) of a network device (200) for the transmission of Ethernet data frames via a two-wire bus (102) shared by several network devices, comprising: - Sending (402) a first arbitration message at a first data rate, which allows the resolution of collisions between two network devices that are transmitting at the same time, the first arbitration message containing information about a priority of a message to be sent by the network device, - Monitoring (404), while the first arbitration message is being sent, whether at the same time a second network device (201) connected to the shared two-wire bus (102) is sending a corresponding second arbitration message, the control method also comprising if the priority of the message to be sent to the second network device (201) is lower than that of the own message to be sent: - Sending (406) the message with the first data rate or with a second data rate which is higher than that than the first data rate, the control method also comprising when the priority of the message to be sent by the second network device (201) is higher than that of your own message to be sent: - Monitoring (408) the two-wire bus in order to detect the end of a transmission of a message sent with a higher priority. Control method (400) Claim 9 wherein the sending (402) of the arbitration message occurs only when either another network device (201) is sending an arbitration message over the shared two-wire bus, or when no network device (201) is sending a message. Computer program product, comprising instructions which, when the program is executed by a microprocessor, cause the latter, the method according to one of the Claims 9 or 10 to execute. Computer-readable data carrier on which the computer program product is after Claim 11 is stored. Vehicle with two or more network devices communicatively connected via a shared two-wire bus Claim 7 .

Images