Network with several nodes and at least one network hub

0001The invention relates to a network with several network nodes. Such networks can be used for example in motor vehicles, in industrial automation (eg sensor systems) and home automation (eg lighting technology, alarm systems, heating systems, air conditioning technology, etc.).

0002In such a network for automotive technology, for example, from the magazine<img file="EP1085705A2_D0001.tif" />Elektronik ", No. 14, 1999, pages 36 to 43 (Dr. Stefan Polenda, Georg Kroiss: <img file="EP1085705A2_D0002.tif" />TTP: <img file="EP1085705A2_D0003.tif" />Drive by Wire "within reach") known TTP protocol (TTP = Time-Triggered Protocol) can be used. This protocol enables secure data transmission and can therefore also be used in networks for safety-relevant devices (eg brakes). In the mentioned article is mentioned as a network structure, a bus system.

0003The invention has for its object to provide a different network with multiple Nerzknoten.

0004The object is achieved by a network of the type mentioned above in that at least part of the network nodes are coupled directly to one another via at least one star node,<ul id="ul0001" list-style="none" compact="compact"><li>that the star node contains a plurality of star interfaces that are assigned to at least one network node, and</li><li>a star interface is provided in each case as a function of a pilot signal for forwarding a message from the assigned network node to the other star interfaces or from another star interface to at least one of the assigned network nodes.</li></ul>

0005The idea underlying this invention is the control of a star node by means of a pilot signal generated by the network nodes. The pilot signal is changed before and after a message transmission so that detects a contained in the star node and a network node associated star interface when the associated network node changes the pilot signal. In the star node, the other star interfaces are then switched to receive only a message from the star interface that receives a message from the associated hub.

0006By means of the star interfaces, a correct line termination at the input of a star interface is achieved and the pilot signal allows unambiguous control of the message traffic.

0007As claim 2 shows, each network node in the network is assigned a specific, periodically repeating period of time for the transmission of its messages. The pilot signal generator in a network node signals when sending a pilot signal this period. The pilot signal can be differentiated from the actual message in various ways. For example, the pilot signal uses a different frequency range than the signal with the message to be transmitted.

0008Claim 3 indicates that each star interface includes first and second switching elements and a pilot signal detector. The switching elements, which can be a switchable amplifier as described in claim 3, control the message flow in a star interface as a function of the pilot signal.

0009A star interface generates during the message transmission, which is displayed by the pilot signal, an enable signal that is ORed to the other star interfaces, as claim 5 describes. Claim 6 indicates that the OR operation can be designed as an OR gate or as a wired-OR connection.

0010The use of a pilot signal to control message traffic allows the selective addition of redundant line connections such that a further star interface is added for each additional line connection. In addition, for each additional line connection, the network node is extended by a respective circuit component for transmitting a message combined with a pilot signal and for receiving such a message. To find the fault and to select a line connection, a control unit is required.

0011A line connection can be a line pair, for example, in the case of symmetrical signal transmission.

0012The invention also relates to a network node which is characterized<ul id="ul0002" list-style="none" compact="compact"><li>the network node is provided for coupling to further network nodes via at least one star node and</li><li>the network node is provided for indicating a message transmission to a star interface of the star node with a pilot signal.</li></ul>

0013Furthermore, the invention relates to a star node in a network for coupling a plurality of network nodes having a plurality of star interfaces, which are assigned to at least one network node, and in each case in response to a pilot signal for forwarding a message from the associated network node to the other star interfaces or from one another star interface to at least one of the associated network nodes are provided.

0014Embodiments of the invention are explained below with reference to FIG. Closer. Show it:<dl id="dl0001" compact="compact"><dt>Fig. 1</dt><dd>a star-structured network with multiple network nodes coupled via an active star node,</dd><dt>Fig. 2</dt><dd>a first embodiment of a star interface in a star node,</dd><dt>Fig. 3</dt><dd>a first embodiment of a first star node,</dd><dt>Fig. 4</dt><dd>a part of a network node with a pilot signal generator,</dd><dt>Fig. 5</dt><dd>a part of a time course of two signals in the network,</dd><dt>6 and 7</dt><dd>two further embodiments of the network,</dd><dt>Fig. 8</dt><dd>a second embodiment of a star interface in a star node,</dd><dt>Fig. 9</dt><dd>a second embodiment of a first star node,</dd><dt>Fig. 10</dt><dd>an embodiment of the network with additional redundant components and</dd><dt>Fig. 11</dt><dd>another embodiment of a part of a network node with two pilot signal generators, detectors and a control unit.</dd></dl>

0015An embodiment of the network according to the invention is shown in FIG. 1 shown. This network includes, for example, four network nodes 1 to 4, which are each coupled via twisted, provided for a symmetrical signal transmission line pairs 5 to 8 (twisted-pair) with each other via an active star node 9. The active star node 9 carries out a line adaptation, so that the line pairs 5 to 8 in the active star node 9 are terminated with the characteristic impedance, and detects a pilot signal transmitted by a network node 1 to 4. If the line pairs 5 to 8 were connected to each other without an active star node 9, a mismatch would result for each pair of lines in the star point due to the impedance discontinuity of Z.₀ to 1/3 Z₀, which is caused by the parallel connection of the other line pairs. To generate a pilot signal, each network node 1 to 4 also contains a pilot signal generator.

0016The active star node 9 contains a star interface for each line pair 5 to 8, which allows a forwarding of the messages of a transmitting node to all other network nodes connected to the active star. An embodiment of such a star interface is shown in FIG. 2 shown. A pair of wires 5 to 8 are connected to the inputs of a switchable amplifier 10, to the outputs of a further switchable amplifier 11, to a pilot signal detector 12 and to a terminating resistor 13. The value of the terminating resistor 13 corresponds to the characteristic impedance and thus serves for correct line termination. When the pilot signal detector 12 detects a pilot signal, it generates an enable signal which is provided to a switching input 14 of the switchable amplifier 10, to an inverting input of an AND gate 15 and via an amplifier 17 and a line 18 to an OR gate 23 (FIG. 3) is guided. When the switchable amplifier 10 is enabled, it provides data to a data line 19 leading to a node. Data from the other star interfaces are also received by this data line 19 and routed via the switchable amplifier 11 to the assigned line pair. Via a line 20 is still the non-inverting input of the AND gate 15, an enable signal via the OR gate 23 (Fig. 3) from another star interface. The non-inverting output of the AND gate 15 is connected to a switching input 21 of the switchable amplifier 11 and via an inverter 16 to an enable input 22 of the pilot signal detector.

0017A switchable amplifier 10 or 11 may also be implemented as a series circuit of amplifier and switch (switching element). In the closed state of this switch, the output signal of the amplifier is forwarded.

0018The pilot signal detector checks whether the associated network node by sending the pilot signal shows that he or she messages Data is transferred. If this is the case, the amplifier 11 is deactivated (in general, he will already be in this state) and the amplifier 10 is activated or Approved. The message signal coming from the assigned line pair is routed to line 19 and thus passed on to the other star interfaces. In addition, the enable signal generated by the pilot signal detector 12 signals the other star interfaces to in turn enable or enable their amplifier 11 to relay the messages to the respective associated network nodes.

0019As shown in FIG. 3, the lines 18 of all the star interfaces in the OR gate 23 are linked. Further, Fig. 3 shows four star interfaces 24 to 27, which are each coupled to the cable pairs 5 to 8. The output of OR gate 23 is connected to lines 20 (Figure 2) of each star interface 24-27. The lines 19 (FIG. 2) of each star interface 24 to 27 are linked to each other via a circuit node 28.

0020In order to prevent a signal coming from a line pair 5 to 8 being fed back into it, the amplifier 11 is deactivated or switched off via the inverting input of the AND gate 15. On the other hand, to prevent the pilot signal detector associated with another network node from in turn activating its amplifier 10, this is done via the line 20 (FIG. 2) uses the transmitted enable signal through AND gate 15 and inverter 16 to suppress the enable signal for the pilot signal detector 12. Thus, the amplifier 10 is switched off or deactivated via its switching input 14. However, the pilot signal detector 12 whose pilot signal causes the associated star interface to receive or receive data remains active to be able to detect the end of data transmission.

0021For proper operation of the active star node 9, it is necessary that the network nodes 1 to 4 send their messages non-overlapping in time. In addition, it must be ensured that for the duration of a certain dead time no network node is active or messages or Sends data. In this state, the star node is switched completely opaque (ie all amplifiers 10 and 11 are deactivated). In this state, a star interface 24 to 27 in the star node 9 waits for a new pilot signal indicating a transmission request of messages.

0022Basically, the pilot signal always has to be sent out before the beginning of the actual message transmission. Only then is it ensured that the active star node 9 is configured in good time and that the beginning of the message reaches all other network nodes.

0023In Fig. 4 It is shown how the pilot signal is generated in a network node 1 to 4 and transmitted via a line pair 5 to 8. When a network node wants to send a message or data to other network nodes, a pilot signal generator 29 receives via a line 30, for example, a start signal. The pilot signal generator 29 then supplies a pilot signal to a multiplexer 31, which is still supplied by a line 32 to be sent data. The signal output by the multiplexer 31 is applied to the associated line pair via an amplifier 33. A signal originating from another network node is passed from the line pair via an amplifier 34 to a line 35 for further processing.

0024The in Fig. 4 shown multiplexer can be designed both as a temporal multiplexer (sending the pilot signal as a start and stop signal respectively before and after the actual message) or as a frequency multiplexer. This means that the pilot signal either as a continuous signal can accompany the entire message to be transmitted or that it can be sent in the form of a start and stop signal. It can eg be ensured by different duration that the start and stop signal are sufficiently different and the change between transmission period and transmission break is not confused.

0025The pilot signal can be generated in various ways. One possibility is that it may be a periodic signal whose frequency range is outside the frequency range used to transmit the messages. This frequency range can be above or below the useful frequency band, but also with a corresponding specification of the useful band due to the narrowband nature of the pilot signal in<img file="EP1085705A2_D0004.tif" />Another possibility is to transmit the actual message as a balanced push-pull signal and the pilot signal as a common-mode signal.The common-mode signal can be designed in the form of both a constant voltage and a periodic signal.A third option for the pilot signal is in that this special message is implemented in the form of the message transmission, upstream and downstream.

0026The proposed invention is particularly well suited for use in networks that work according to the TTP protocol for real-time communication, for example in the motor vehicle (see Electronics Issue 14/1999: <img file="EP1085705A2_D0005.tif" />TTP: <img file="EP1085705A2_D0006.tif" />Drive by Wire "within reach", pages 36 to 43). In this protocol, on the one hand, it is defined when which transmitter is allowed to transmit using the TDMA (Time Division Multiple Access) access method, and, on the other hand, an idle time (interframe gap) is defined in which no transmitter is allowed to transmit. Through this mechanism, it is immediately ensured that the active star node 9 always returns to the idle state. It is thus ensured with the TDMA method that only one network node may send a message at a predetermined time and, by means of the pilot signal transmitted by it in the star node, activates or initiates the star interface assigned to it for forwarding messages.

0027An additional advantage is that for controlling the so-called bus guardian (bus guardian), a control signal must be present in the Nerzwerkknoten, which is shortly before the start of the message transmission. This control signal can be used directly to drive the pilot signal generator 29 by this control signal is supplied on the line 30 to the pilot signal generator 29.

0028In Fig. 5 this control signal is marked with BG and the actual message with data. The control signal BG is, for example, in a low state during the transmission of a message. During this low state of the control signal, the message should be transmitted. A first period T1 after a change of the control signal in the low state and a second period T2 before a change of the control signal in the high state must be chosen so that the active star node 9 is configured correctly and remains to transmit a message without error can. It remains to be noted that the TTP protocol supports different (constant) message delay times between different network nodes in the network. Thus, caused by the active star node 9 delay time does not violate the TTP protocol.

0029The network according to the invention makes it possible to transmit a pilot signal with any type of signal transmission for the messages from a network node 1 to 4. For example, a symmetrical push-pull transmission, single-conductor transmission or carrier-frequency-modulated transmission can be selected for the message transmission. With common mode coupling of the line pairs 5 to 8, a supply voltage could possibly also be transmitted with the message transmission.

0030Another embodiment of a network is shown in FIG. 6 shown. This network is identical to the one shown in FIG. 1 shown network (mink nodes 1 to 4 and line pairs 5 to 8) almost identical except for four additional network nodes 36 to 39. The network node 36 is coupled directly to the active star node 9 via another line pair 40. The network node 37 is connected to the line pair 5 via a line pair 41. The network node 38 is connected via a line pair 42 to the line pair 8 and the network node 39 via a line pair 43 to the line pair 42.

0031The exemplary embodiment shown in FIG. 7 again shows the topology of the network known from FIG. 1 with the network nodes 1 to 4, the cable pairs 5 to 8 and the active star node 9. Via a pair of lines 44 the active star node 9 is connected to another active one Star node 45 coupled to the turn more nodes can be connected. By way of example, a network node 46 is shown here.

0032Instead of the in Fig. 3 required OR gate 23, this OR operation can also be realized directly by a wired-OR operation. A star interface suitable for the wired-OR operation is shown in FIG. 8th. This star interface according to FIG. 8th is identical to the star interface of FIG. Second In the Fig. 8th The output of the amplifier 17 does not lead to the line 18, but to the line 20 (Wired-OR). Thus, the output of the amplifier 17 is connected to the inverting input of the AND gate 15. The amplifier 17 is shown in FIG. 2 when <img file="EP1085705A2_D0007.tif" />Push-pull "amplifier and in Fig. 8 as <img file="EP1085705A2_D0008.tif" />Open-Collector "- resp. <img file="EP1085705A2_D0009.tif" />Realized open-drain "amplifiers.

0033This Wired-OR connection reduces the wiring complexity and eliminates the OR gate 23 in the active star node 9 and results in a simple expandability of the network with other network nodes by eliminating the otherwise available in different variants OR gate. The star interfaces 24 to 27 are in this case connected to their respective leads 19 and 20, so that, as shown in FIG. 9 shown forming two circuit nodes 47 and 48. In addition, only one resistor 49 is provided, which is coupled on the one hand to the circuit node 47 and on the other hand to a supply voltage. This resistor, together with the amplifiers 17 of each star interface, forms the wired OR combination.

0034To increase the availability of network nodes in a network, as shown in FIG. 10 shown, several pairs of lines and thus the associated circuit component for generating a pilot signal (see. Fig.4) doubled. The network of FIG. 10 includes four network nodes 50 to 53 and a star node 54. The network nodes 50 and 51 are each coupled via a line pair 56 and 57 and the network nodes 52 and 53 via two pairs of lines 58 to 61 to the star node 54. This means that the network node 52 is connected to the star node 54 via the line pair 58 and 59 and the network node 53 via the line pair 60 and 61.

0035In Fig. 11 FIG. 4 illustrates how a pilot signal is generated in a network node 52 and 53 and transmitted over one of the line pairs 64 or 65. The pilot signal generators 66 and 67 serve to generate a pilot signal. By supplying a control signal on lines 68 or 69 from a controller 82, it is determined which pilot signal generator (66 or 67) is generating a pilot signal. The pilot signal generator 66 or 67 provides such a pilot signal to an associated multiplexer 70 or 71st The multiplexer 70 or 71 still receives a message or data via a line 72 or 73rd The output signal of the multiplexer 70 or 71 is via a switchable amplifier 74 or 75 the respective line pair 64 or 65 fed.

0036Data originating from the star node 54 is fed via the pair of lines 64 and 65 to an amplifier 76 and 77, respectively. The amplifier 76 or 77 supplies the data via a line 78 or 79 to further circuit components of the network node.

0037With the output of each amplifier 76 and 77 is a pilot signal detector 80 or 81 connected. If a pilot signal has been detected by a pilot signal detector 80 or 81, this is communicated to a control unit 82 (state machine). This controller 82 controls the two switchable amplifiers 74 and 75. Further, the controller 82 controls the pilot signal generators 66 and 67 via the lines 68 and 69. From the network node, the already mentioned above start signal is passed on a line 83 to the control unit 82, which therefrom a control signal for the pilot signal generators 66 and 67 forms. Via a line 84, the control unit 82 can report states to the network node, for example.

0038The circuit elements 66, 68, 70, 72, 74, 76, 78 and 80 form a first circuit component 85 and the circuit elements 67, 69, 71, 73, 75, 77, 79 and 81 form a second circuit component 86. The circuit components 85 and 86 are each constructed identically.

0039This control unit 82 manages the line pairs 64 and 65 connected to the network node in such a way that the functionality of the line pairs 64 and 65 and of the circuit components 85 and 86 is constantly checked.

0040During message reception, the control unit 82 checks whether the pilot signal actually arrives over both line pairs 64 and 65. The arrival of a pilot signal on one of the two redundant line pairs signals the beginning of a message reception. The control unit 82 checks during a certain period, whether after the arrival of the first pilot signal on one of the line pairs 64 or 65 the pilot signal has arrived on the other line pair. If this is not the case, it can be assumed that the other line pair is faulty.

0041Sending is basically only on one of the redundant line pairs. In order to achieve that even if a circuit component 85 or 86 the communication can be maintained, it is checked at each transmission, whether in fact the pilot signal accompanying the message is sent out. The structure of the star node ensures that the pilot signal is returned to the transmitting network node via the line pair that is currently not being used for transmitting a message. Here too, the control unit 82 checks whether the pilot signal arrives on the other line pair within another specific time interval after the start of the transmission of the message.

0042The combination of these two tests makes it possible to detect a defect of the currently used for the transmission of messages Shock component 85 and 86 and the line pairs 64 and 65 and to switch the intact circuit component with the associated intact line pair of the control unit 82.

0043If there is no error and control unit 82 has registered both line pairs 64 and 65 and the two circuit components 85 and 86 as operative, circuit components 85 and 86, for example, will alternately transmit to the associated star interface. In the event of a fault, only the circuit component 85 or 86 registered as error-free transmits.