• News
• About
• Products
• Services
• Training
• Partners
• Contacts

CAN/Ethernet/Profibus Bridge

The developed protocol can be used as a isolation layer on top of several network mediums applied in building distributed control systems. Such network mediums are: CAN, Ethernet, Profibus. The features of the protocol are:

• Support for unified connection with the lower level protocols (MAC level protocols) while using different types of communication interfaces (CAN interface with/without communication processor, Ethernet interfaces, Profibus interfaces).
• Registering of users' tasks, which are sinks for messages
• Filtering of incoming messages.
• Passing received messages toward registered user tasks.
• Load balancing of outgoing messages between the established network interfaces, this is possible both over interfaces of the same type and over interfaces with different network types.
• Balancing of network interface load inorder to avoid any network congestions, when such conditions are detected the excess traffic can be sent over another medium (if the appropriate connections have been created), which would allow the system to continue functioning.
• Guarantee of system schedulability.

With the implemented infrastructure its possible to create distributed control systems, using different communication mediums (CAN, Ethernet, Profibus) in different network segments. During implementation phase of the tasks running on the nodes in the distributed control system its required that these tasks are designed against a set of simple rules, and later using the proposed architecture the tasks can use different network mediums (depending on what network is available or applicable). At later stages in the product the communication medum can be interchanged without affecting system stability (as long as the new communication medium supports the initially designed network speeds required for stable system communication). When increased network communication stability is required it can be achieved by using parallel branches in the network, with the proposed architecture the parallel links can be kept as backup links and in case of failure of currently used links the backup links can be brought online without significant delays, thus the system can continue to operations. (The backup links can be from any of the supported medium types, eg. the main channel is Profibus while the backup links is CAN or Ethernet). In case higher throughput of the network is required the network architecture can be built using packet load-balancing, when several parallel links are established and configured they can be used in parallel and thus decreasing the load on each of the links and increasing network throughput. When failure is detected in one of the used links that link is brought down without further causing any message delivery problems, messages are delivered to each of the receivers and no message loss is allowed.