WO2009098616A1

WO2009098616A1 - Ring topology, ring controller and method

Info

Publication number: WO2009098616A1
Application number: PCT/IB2009/050359
Authority: WO
Inventors: Andries Van Wageningen; Joern Ungermann; Peter Fuhrmann
Original assignee: Nxp B.V.
Priority date: 2008-02-06
Filing date: 2009-01-29
Publication date: 2009-08-13

Abstract

The invention relates to a ring topology (1) of nodes (2,3,4,5) wherein a node contains at least one a host (6,7,8,9) and a ring controller (10,11,12,13) with two interfaces and a predetermined amount of nodes are connected by means of a data connection between respective interfaces of two different nodes, and all nodes are connected to a ring structure and the connection between the interfaces is a FlexRay connection.

Description

RING TOPOLOGY, RING CONTROLLER AND METHOD

FIELD OF THE INVENTION

The invention relates to a ring topology as claimed in claim 1 , a ring controller as claimed in claim 8 and method to control a data transmission within a ring topology as claimed in claim 12.

BACKGROUND OF THE INVENTION

Automotive vehicles are using a large number of electronic units to control the actuation or performance of actuators or components of the vehicle. Such electronic units are usually connected to a communication network. Such dependable automotive communication networks typically rely on time-triggered communication protocols like TTP/C (Time Triggered Protocol Class C) or FlexRay, based on broadcast data or messages according to a pre-determined TDMA (Time Division Multiple Access) scheme. A typical fault-tolerant time-triggered network consists of two or more communication channels, to which different nodes are connected. Such nodes are typically control units of electronically controlled devices. Each of those nodes consists e.g. of bus drivers, a communication controller and/or eventually a bus guardian device for each bus driver and the application host. The bus driver creates the connection to the data channel and writes the information to be sent via the data bus to the channel and it reads the signals from other nodes or units from the channel.

Different topologies are possible to connect the nodes via these channels. Most favourite is the bus topology, because it limits the number of required connectors and bus drivers. However due to its shared character, the number of nodes and/or the maximum speed on the bus might be limited to guarantee the signal quality on the bus. Another commonly used topology is the star-topology. By using an active star-coupler the signal quality on the communication line is improved, compared the situation where nodes are connected via a passive bus. An active star-coupler allows connecting more nodes in a single cluster than a passive bus. A conventional star-coupler works on physical level forwarding data from one selected input port to all output ports at a time. On protocol level, it does not show a difference between a bus and a star topology. In a hybrid topology, both bus- and star-topologies can be combined in one network. For example a central star-coupler can connect not only a number of nodes on individual base, but also small sub-networks, each containing a number of nodes connected to a bus. The FlexRay protocol as been as today is specified for a speed of 10 Mbit/s and supports bus-, star- and hybrid topologies as mentioned above. The FlexRay protocol additionally allows other protocols on top of it. Messages with their own identifiers for example could be applied to allow the identification of additional services not addressed by the FlexRay protocol it selves.

With regard to the increasing amount of transmitted data and the increasing amount of connected units, future applications might require a higher bandwidth as is offered by the FlexRay technology today. For higher bandwidth or bit rates, e.g., 100 Mbit/s, the bus topology does not fit very well due to multiple reflections on the bus and different distances between the nodes connected to the bus.

One solution will be to limit the possible topologies, to active star-coupled only. This will lead to the situation whereby each node is individually connected to the active-star, resulting in the usage of point-to-point connections on physical level. The disadvantages of this solution are the following compared to the bus-topology that a higher total cable length is required with in the vehicle or within the application and that twice the number of bus-drivers and connectors is required. Since the cabling and bus-driver costs are significant factors in a vehicle with many electronic components or units, these disadvantages are highly relevant. For applications that require redundant communication channels, these disadvantages become even more relevant since the number of bus-drivers and the cable length will at least be doubled. A network with N nodes, each connected to two active star- couplers requires 4N connectors and bus-drivers.

Compared to the star-topology, a ring-topology requires less cabling. A network with N nodes connected in a ring requires 2N connectors and bus-drivers. Because each node can reach any other node in two directions, it has redundant interconnection of the nodes by nature.

The FlexRay protocol however is not specified for ring-topologies. One of the problems would be the synchronization of the network, because each node will cause an additional delay in the transfer of synchronization frames in the network.

OBJECT AND SUMMARY OF THE INVENTION

It is an object of the invention to create a method that allows the application of the FlexRay protocol in a ring-topology. The object of the invention will be solved with respect to the ring topology according to the features of claim 1. The inventive ring topology of nodes comprises a node which contains at least one host and a ring controller with two interfaces and a arbitrary amount of nodes are connected by means of a data connection between respective interfaces of two different nodes, and all nodes are connected to a ring structure and the connection between the interfaces is a FlexRay connection.

According to the invention it is of advantage that the ring controller of a node is able to transmit data via an interface to another interface of another node by means of the FlexRay connection. Furthermore it is of advantage that the ring controller of a node is able to receive data from an interface from another interface of another node by means of the

FlexRay connection and additionally it is of advantage that the ring controller of a node is able to copy data received from an interface to the host of the node. Furthermore it is of advantage that the ring controller of a node is able to forward data received from an interface (24,25) to another interface.

Furthermore according to one embodiment of the invention the ring controller of the last receiving node terminates the circulation of a data message.

According to another embodiment of the invention the first transmitting ring controller of a message terminates the transmission of the message after receiving the message again. Furthermore it is of advantage that the ring controller of a predetermined node terminates the forwarding of a data message in the ring, whereby for each data message a node is assigned to terminate the message forwarding.

Additionally it is of advantage according to an other embodiment of the invention that the ring controller that inserted a message terminates the forwarding of the message after receiving the same message circulated in the ring.

The object of the invention concerning the ring controller will be solved using a ring controller for the use within the above mentioned ring topology characterised in that the controller contains two interfaces and each interface is connected to a bus driver, the bus driver is connected to a protocol engine and the protocol engine is connected to a controller host interface which is extended by an function that is able to recognize if a received message is meant for forwarding in order to put it into a message buffer determined for forwarding the message, wherein the controller host interface is connected to a message buffer.

The object of the invention concerning the method will be solved by the method to control a data transmission within a ring topology of nodes wherein a node contains at least one a host and a ring controller with two interfaces and wherein an arbitrary amount of nodes are connected by means of a data connection between respective interfaces of two different nodes, and all nodes are connected to a ring structure and the connection between the interfaces is a FlexRay connection, wherein the ring controller of a node is able to transmit data via an interface to another interface of another node by means of the FlexRay connection and/or the ring controller of a node is able to receive data from an interface from another interface of another node by means of the FlexRay connection and/or the ring controller of a node is able to copy data received from an interface to the host of the node.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other features and advantages of the invention will be apparent from the following description of an exemplary embodiment of the invention with reference to the accompanying drawings, in which:

Fig. 1 shows a schematic diagram of an inventive ring topology;

Fig. 2 shows a diagram of the function of a ring controller; Fig. 3 shows a diagram of the function of a ring controller;

Fig. 4 shows a diagram of the function of a ring controller;

Fig. 5 shows a diagram of the function of a ring controller;

Fig. 6 shows a diagram of a ring-topology of a number of nodes or clusters;

Fig. 7 shows a diagram of a ring-topology of a number of nodes or clusters; Fig. 8 shows a node with host and ring controller; and

Fig. 9 shows a node with host and ring controller.

DESCRIPTION OF EMBODIMENTS

The invention described below creates a method or a mechanism that allows the application of the FlexRay protocol in a ring-topology. In order not to offend the existing protocol, a ring topology can be build by using every node as a two-way gateway between two FlexRay sub-networks; a separate FlexRay network is used as link between every two nodes connected in the ring topology. To establish connections between nodes that are not directly connected to each other, a new protocol is added on top of the existing FlexRay protocol. This new protocol is based on the store-and-forward principle. It defines the handling of messages in each node. These messages can be freely allocated to different slots at each FlexRay interface. The protocol defines the insertion, extraction, forwarding and termination of the messages. The invention is described starting from an abstract definition of the required communication functions for a node in the ring-topology, followed by an overview of the building blocks inside a node and finally the mapping of the functions to these building blocks.

Figure 1 shows an example of an inventive ring topology 1 for four FlexRay nodes 2, 3, 4, 5. The mentioned number of nodes is only an example and not limited. These nodes 2, 3, 4, 5 are containing respectively a host 6, 7, 8, 9 and a ring controller 10, 11, 12, 13 and are connected with two other nodes via the ring controller 10, 11, 12 and 13 by means of two FlexRay connections or interfaces FlexRay-1 to FlexRay-4 14, 15, 16, 17. Each FlexRay connection or interface 14, 15, 16, 17 allows transmitting and receiving frames. The inventive embodiment contains four independent FlexRay clusters, whereby each FlexRay cluster connects two nodes 2, 3, 4, 5. Each FlexRay cluster has its own schedule, but preferably the schedules are identical to realize symmetrical performance in the ring. The following Figures 2 to 5 demonstrate the basic communication functions for a node as shown in figure 1 and which are labelled with the references 2, 3, 4, 5.

Fig. 2 shows the insertion functions of a ring controller 20. If the host 21 transmits a message 22 towards the other nodes, the ring controller 20 can send this message 22 in two different directions 23. The direction 23 is determined by the configuration of the ring controller 20 e.g. by a message address lookup table. Transmission of a message 22 in one direction 23 will be sufficient for non-redundant communication. In this case the configuration can be optimised by choosing the direction having the shortest number of FlexRay clusters to the destination hosts. Insertion of the message on both FlexRay interfaces supports the realization of redundant communication. Fig. 3 shows the extraction functions of the ring controller 20. A message 22 is received by one of the two interfaces 24 of the ring and offered to the host 21 for reception. Message reception can be realised by message filtering techniques, e.g. with a message address lookup table. A node can extract a message 22 from one of the interfaces 24 for non- redundant communication and from both interfaces for redundant communication Fig. 4 shows the forwarding functions of the ring controller 20. The ring controller 20 forwards a message 22 received from one of the two interfaces 24 to the other interface 25 without forwarding the message to the host. The function can be realised in two different directions. For non-redundant communication one direction will be sufficient. For redundant communication both directions are needed in parallel. The configuration of the forwarding functions can be realised with message address tables. Fig. 5 shows a combination of extraction and forwarding functions. The ring controller 20 receiving a message 22 on one of the two interfaces 24 forwards it to the other interface 25 and in parallel offers e.g. a copy 26 of this message 22 to the host 21. Again this can be realised in one direction in the ring for non-redundant communication, but also in two directions for redundant communication.

Fig. 6 shows a diagram 30 of a ring topology of a network 31 of four nodes 2, 3, 4, 5 and the usage of the above described functions for a non-redundant communication embodiment. Every node may represent a cluster of nodes, as explained above. Node 1, 2 sends a message 36 to node 2, 3 and to node 3, 4. The ring controller 1, 10 inserts the message 36 on its interface 38 connected to the FlexRay 1 connection 14. The ring controller 2, 11 extracts the message from its FlexRay 1 interface 41 and forwards it to its FlexRay 2 interface 42. Furthermore the ring controller 2, 11 offers the message 36 or a copy of the message to host 2 of node 2, 3. The ring controller 2, 11 transfers the message to the interface 42 and transmits the message to the FlexRay 2 connection 15. The ring controller 3, 12 receives the message via the interface 45 and the ring controller 3, 12 extracts the message from its FlexRay 2 interface 45 and offers it to host 3, 8. Since it does not forward the message, ring controller 3, 12 terminates the transmission of the message on the ring.

As in Fig. 6, the ring 48 can be configured such that the ring controller that is last in the chain of destinations for this message terminates the message on the ring 48. This will result in efficient usage of the available bandwidth in the ring 48, but requires dedicated configuration of each of the nodes 2, 3, 4, 5. In case a new ring controller is added to the ring 48, or an existing one is removed from the ring 48 re-configuration requires quite some overhead.

A more flexible configuration method for regarding re-configuration is shown in Fig. 7. The ring controller (RCl) 10 that inserts a message 36 into the ring 48 also terminates the message after the message 36 has travelled through the complete ring 48 and has been received by the sending ring controller 10 again. All other ring controllers (RC2, RC3, RC 4) of the other nodes node 2, node 3 and node 4 forward this message 36 and if needed offer the message or a copy of the message to the respective hosts host2, host 3, host 4.

Redundant communication can be realised by applying this method for both directions on the ring 48. In this case RCl inserts the message on both its interfaces 38, 49, and the ring controllers RC 2, RC 3 and RC 4 are forwarding the message 36 in both directions, and after travelling trough the complete ring, the ring controller RCl, 10 extracts the message 36 on both its interfaces 38, 49 and terminates the travelling of the message 36 on the ring 48.

Fig. 8 shows an embodiment of the structure 50 of a ring controller 51 in a node 52 of the network. The FlexRay interfaces 53, 54 are interfaces, each containing a bus driver (BD) 55, 56, a protocol engine (PE) 57, 58 and a Controller Host Interface (CHI) 59, 60. For the purpose of message insertion, extraction and forwarding, a Controller Host Interface is extended (Ext) 61,62 with extension 61,62 with extra functionality. This function of the extension is able to interpret the address field of the message. Communication between Controller Host Interface 59, 60 and the host 63 is arranged via message buffers 64, 65. These buffers 64, 65 can be realised separately e.g. with two memory blocks for each of the two FlexRay interfaces 53, 54, but they can also be realised with a shared facility with e.g. only one memory block for both FlexRay interfaces. The forwarding function in this embodiment of the invention is located in the ring controller. It can be realised in different ways. One method is to place the forwarding function FW, 66 between reception forwarding buffers and transmission forwarding buffers 64, 65. A message extracted from one FlexRay interface by the Controller Host Interface CHI 59, 60 is stored in a reception forwarding buffer 64, 65. From there it is copied by the forwarding function 66 into the transmission forwarding buffer 64, 65 of the other FlexRay interface. In case the message needs to be offered to the host, the Controller Host Interface CHI 59, 60 copies it into a host reception buffer as well.

Figure 9 shows an alternative method. It shows that the forward buffer 70 is placed between the two Controller Host Interfaces 71, 72. This could be realised by applying a shared memory. A message extracted from one FlexRay interface is stored in a forwarding buffer 70 by the Controller Host Interface CHI 71, 72. From there it is read by the second Controller Host Interface CHI 72, 71 to insert into the second FlexRay interface. With regard to FlexRay subsystems each point-to-point connection within a FlexRay ring might be realised with a two-node FlexRay cluster. This might prevent failures of a single node from affecting the overall ring structure. Furthermore it is possible to use the FlexRay protocol either in a one-channel or a two-channel configuration. The second channel can be used to increase the bandwidth, the data integrity and/or the link reliability. If the same message is inserted on both channels and handled identically by all nodes then the data integrity is increased since the receiving nodes can compare the message received on both channels and compare it for deviations. If different messages are being sent on both channels, the bandwidth is increased. The second channel can also be used for link reliability, by sending the same message in one direction on the first channel and in another direction on a second channel. Since FlexRay is only a half- duplex protocol, this reduces the propagation delay of messages by restricting the direction within which messages can travel on a ring. The FlexRay feature of hardware supported 'message IDs' might be used to identify the messages, which is normally done via the FlexRay slot ID. This enables e.g. an easy implementation of the extraction function, where message IDs recognition can be implemented in the ring controller without bothering the host. Each message ID is uniquely given to a certain message type containing specific signals and only one node has the right to insert a message with said message ID. Any node 'interested' in this message can recognise the message by its 'message ID' and offer the contained data to its application.

The FlexRay cycle might contain two static slots, where each node in the cluster is assigned one of them. According to the invention it is proposed to use two slots to exchange data or network information or management information, e.g., to slowly synchronise all the FlexRay clusters making up the ring.

References

1 ring topology

2 node

3 node

4 node

5 node

6 host

7 host

8 host

9 host

10 ring controller

11 ring controller

12 ring controller

13 ring controller

14 FlexRay interface, communication

15 FlexRay interface, communication

16 FlexRay interface, communication

17 FlexRay interface, communication

20 ring controller

21 host

22 message

23 direction

24 interface

25 interface

26 copy

30 diagram

31 network

36 message 38 B interface

41 FlexRay interface

42 FlexRay interface

5 FlexRay interface

48 ring

49 interface - Ii

50 structure

51 ring controller

52 node

53 FlexRay interface

54 FlexRay interface

55 bus driver

56 bus driver

57 protocol engine

58 protocol engine

59 controller host interface

60 controller host interface

61 extension

62 extension

63 host

64 buffer

65 buffer

66 forwarding function/buffer

70 forward buffer

71 controller host interface

72 controller host interface

Claims

CLAIMS:

1. Ring topology (1) of nodes (2, 3, 4, 5) wherein a node (2, 3, 4, 5) contains at least one host (6, 7, 8, 9) and a ring controller (10, 11, 12, 13) with two interfaces (24, 25) and a arbitrary amount of nodes are connected by means of a data connection (14,15,16,17) between respective interfaces of two different nodes, and all nodes are connected to a ring structure and the connection (14,15,16,17) between the interfaces is a FlexRay connection.

2. Ring topology according to claim 1, wherein the ring controller (10, 11, 12, 13) of a node is able to transmit data via an interface (24, 25) to an interface (24,25) of another node by means of the FlexRay connection.

3. Ring topology according to claim 1 or 2, wherein the ring controller (10, 11, 12, 13) of a node is able to receive data from an interface (24, 25) from an interface (24, 25) of another node by means of the FlexRay connection.

4. Ring topology according to claims 1, 2 or 3, wherein the ring controller (10,11,12,13) of a node is able to forward data received from an interface (24,25) to another interface (25,24).

5. Ring topology according to claim 1, 2, 3 or 4, wherein the ring controller (10, 11, 12, 13) of a node is able to copy data received from an interface to the host of the node.

6. Ring topology according to one of the preceding claims, wherein the ring controller (10, 11, 12, 13) of a predetermined node terminates the forwarding of a data message in the ring, whereby for each data message a node is assigned to terminate the message forwarding.

7. Ring topology according to one of the claims 1 to 6, wherein the ring controller (10, 11, 12, 13) that inserted a message terminates the forwarding of the message after receiving the same message circulated in the ring.

8. Ring controller (10, 11, 12, 13) for the use within the above mentioned ring topology characterised in that the controller contains two interfaces and each interface is connected to a bus driver, the bus driver is connected to a protocol engine and the protocol engine is connected to a controller host interface which is extended by an function that is able to recognize if a received message is meant for forwarding in order to put it into a message buffer determined for forwarding the message, wherein the controller host interface is connected to a message buffer.

9. Ring controller (10, 11, 12, 13) according to claim 8, wherein the controller host interfaces of the two interfaces are connected via a shared message forward buffer, wherein a controller host interface puts a received message into the message forward buffer from where it is taken by the other controller host interface for transmission.

10. Ring controller according to claim 8, wherein separated message buffers of the two interfaces are connected via a forwarder, wherein the forwarder forwards a message from a message receive buffer to a message transmit buffer.

11. Method to control a data transmission within a ring topology (1) of nodes (2,

3, 4, 5) wherein a node contains at least one a host (6, 7, 8, 9) and a ring controller (10, 11, 12, 13) with two interfaces (24, 25) and wherein an arbitrary amount of nodes are connected by means of a data connection between respective interfaces of two different nodes, and all nodes are connected to a ring structure and the connection between the interfaces is a FlexRay connection, wherein the ring controller of a node is able to transmit data via an interface to another interface of another node by means of the FlexRay connection and/or the ring controller of a node is able to receive data from an interface from another interface of another node by means of the FlexRay connection and/or the ring controller of a node is able to copy data received from an interface to the host of the node and/or the ring controller of a node is able to forward data received from an interface to another interface.

12. Method according to claim 11, wherein a ring controller (10, 11, 12, 13) of a node (2, 3, 4, 5) is able to transmit, receive and forward data in two directions on the ring via its interfaces.