"""Distributed control systems are widely used in today's even more complex technologies. Usually they are built on networks basis, where the nodes belongs to heterogeneous networks. Due to their specific advantages, CAN and LonWorks networks became de facto standard in different, often overlapping areas in industry, science and many others. The common operation of CAN and LonWorks devices involves however sustained high speed communication, transferring data in differing formats, with different data rates and using network's specific communication medias. These particular conditions require an appropriate protocol and physical link converter - a CAN-to-LonWorks gateway, in order to satisfy the requirements of both networks and to combine additional functions to control the direction and the amount of the transferred data, information updating, data integrity etc. A powerful gateway proposed in this paper is designed using In Application reprogrammables PSD components, poviding for flexible functionality, small size, low power consumption and low cost - all the qualities of prime importance. It is structured as two independent CAN and LonWorks nodes, each one running in its own network on peer-to-peer basis. Internally, they communicate via high speed Dual Port RAM, used as postbox, without need of any kind of software or hardware synchronization."""

As Tele-robotics continuously receives industrial attention, an embedded CANopen Master Stack, accessible via the Internet, is proposed. A configurable real-time operating system permits multi-threaded development whilst PLD technology ensures system evolution. The impact thread prioritization has on system performance is investigated, whilst initial design considerations are also presented.

Both CAN- and Ethernet-based networks are widely deployed in millions of nodes and applications world-wide. CAN was initialy developed for automotive and is nowadays very popular in the automation industry, too. For many years, Ethernet based networks are used in automation on higher levels. Due to newly developed real-time protocol extensions, this IT network technology becomes more and more popular on sensor and actuator level as well. CANopen is one of the most extensive selection of industrially applied device and application profiles. So far it was only used on CAN bus systems. With the growing demand for Ethernet on machine level, it was logical to integrate CANopen with Ethernet protocols in general and real-time Ethernet in particular

Distributed embedded systems that require real-time performance need a network capable of deterministic access delay. CAN is one such network that became widespread in recent years due to its electrical robustness, low price, and priority- based access control. However, its use in safety-critical applications has been controversial mainly due to dependability limitations that arise from its bus topology, e.g. the existence of many possible points of failure. In this paper we propose and present a star topology that exhibits improved fault diagnosis and isolation mechanisms with respect to other commercially available hubs. Our hub1 is fully compatible with existing CAN controllers but requires double links. The paper presents a prototype implementation configurable with 4 to 16 ports, describes its architecture and presents some performance results.

"""In this paper we propose a Hardware-in-the-Loop (HIL) approach to implement a preliminary architecture for force-feedback control in steer-by-wire (SBW) applications. A brushless DC motor (BLDC) is used as force feedback actuator. The determination of the position of the BLDC rotor plays a key role in the control algorithm. To obtain a reliable rotor position a classic triple static redundancy (TMR) is implemented. The position signals from the encoder integrated on the motor are computed in three different ways: using a 8-bit microcontroller, a 16-bit microcontroller, and, last, using the software module integrated in a virtual hardware development tool. The virtual hardware platform operates as voter, too. The position is the output of the voting algorithm and it is sent to the 16-bit platform that controls the motor and provides the correct output PWM signals. The communication between virtual hardware and real hardware uses CAN bus. The bus is monitored by a dedicated development tool. Steer-by-wire is a safety critical application and therefore requires time-triggered protocols. In this preliminary architecture a dedicated network has been implemented and therefore the disadvantages of the event-triggered protocol are considerably reduced. Experiments at different baudrates confirm that the voting algorithm produces correct results also in case of failure in one of the modules of the TMR architecture and it is not conditioned by bus loads. This means that the torque control algorithm of the BLDC motor can generate on the steering wheel (directly connected to the motor) a drive feeling like the one produced by a traditional steering system also in this fail- mode."""

A new communications system based on CAN is presented in this work. This system has been implemented to interconnect the different devices that compose a mobile robot, able to operate in environments with elevated electromagnetic noise, being even tolerant to sabotages provoked by means of high electromagnetic fields. The system is based on a shielded communications hub, that interconnects diverse nodes by means of CAN. On the other hand, the point to point communication between each node and the hub is carried out by means of economic optical fiber. With this approach all robot devices are connected by means of CAN, although a optical fiber arrives to each device. In this manner, the total immunity to the electromagnetic noise that the optical fiber transmission presents has been added to all the advantages of a system interconnected by means of CAN. The communications system has been installed in the mobile robot obtaining highly satisfactory experimental results.

The approach to both the architecture and the underlying technologies of a highly distributed systems platform for home and building automation is presented here. This concept is based on the capability to integrate a number of deeply embedded devices that share information among themselves and act as a single system, extracting knowledge from the data collected at the various locations and using this information to intelligently react to events and circumstances in the surrounding area, so as to economically provide comfort, safety and security to people working or living in that controlled environment. Special emphasis is put on the human interfacing to technology, by lifting up the traditional focus from the level of applications to that of services, therefore aiming at serving user's needs, but relieving them from low-level platform concerns. Middleware is used to address the issue of development barriers for services, firstly by providing means for graphical programming in order to rapidly create in-house living scenarios (i.e., complex applications, involving different, interacting variables and conditions) by simply drawing a block diagram that, with recourse to a library of pre-constructed models, addresses the various scattered units and invokes the respective local actions. Units in the system are of two different kinds, as they fit in electrical switchboards (where are concentrated the protective devices to power circuits) or, alternatively, are dispersed throughout the buildings, in relatively larger quantities. The former being integrated over CANbus, and the latter over a wireless cluster tree network based on IEEE 802.15.4 (ZigBee) technology, a unified approach to high-level remote procedure calls was attempted, thus encompassing both wired and wireless networks, which, being based on CANopen high-level protocol, allows a more effective co-operating service-computing platform.

This paper aims at defining an interface for the CAN data-link layer that both fits in well with the IEEE Std 1003.1 socket paradigm, and allows the user to access the full range of capabilities implemented by CAN interface controllers. At the same time, this paper also attempts to set up a set of specifications and guidelines for the actual implemen- tation of the proposed interface.

This paper will discuss the experiences and challenges with the implementation of up- integrated CAN transceivers found on system-basis chips (SBCs) for the automotive market segment. These SBCs exist in an extremely harsh environment, factors such as system interoperatibility, enhanced electrostatic protection, and electromagnetic interference need to be understood and designed in the integrated component to help reduce issues at system level. This is further complicated as the needs for up- integration forces a less than idea silicon process selection to maintain cost goals for the SBC. Detailed discussion of lessons learned include the silicon process development of ESD robust structures at the device level using a lateral-DMOS silicon controller reticifier; the addition of clamp structures to protect the device during short circuit conditions when a CAN choke is used; the influence of fault protection structures on the robustness of the receiver to electromagnetic interference during direct power injection testing; the future market trends for system-on-a-chip development and the impact of process selection to ensure the feasibility of an up- integrated transceiver; and system power up/down issues to minimize bus distrubances. The paper will conclude with future challenges related to up-integrated CAN transceivers.

Interoperability between CANopen devices is universally desired by most CANopen device integrators and other users. To this end, the CiA has defined a CANopen Conformance Test which checks EDS files and the devices for adherence to CiA specifications. However, these checks are against static requirements, implying that 2 devices which pass the Conformance Test may still not be able to actually interoperate. The CANopen protocol deliberately does not prescribe the dynamic timing parameters needed for true device interoperability, as different applications have differing requirements. Recognizing this fact, this paper does not set out to prescribe such specifications. Instead, building on TR 308, this paper describes a standard set of time value measurement definitions and a means for sharing these dynamic values between CANopen devices which provides a user with a means of quickly assessing interoperability & compatibility.

"""The Office of Research and Development of the Federal Railroad Administration (FRA) is sponsoring a revenue service demonstration of Advanced Train Systems featuring new technologies for improving safety and efficiency in freight train operations. The project, which commenced in 1999, is part of the Rolling Stock Program Element in FRA’s Five-Year Strategic Plan for Railroad Research, Development and Demonstrations. The demonstration system, referred to as the On-Board Monitoring and Control System (OBMCS), features an integrated package of sensors and actuators for monitoring and controlling mechanical components on freight trains. The OBMCS includes sensors to monitor bearings, wheels, brakes and trucks and actuators (referred to as advanced components) for remotely controlling parking brakes, angle cocks, cut levers and a cushion unit lockout system to eliminate slack. The demonstration of the OBMCS with advanced components will commence in 2005. An open system architecture based on CAN technology provides the framework for integration and control of the advanced components. The CAN bus network is also employed to monitor bearing temperatures and the status of brake piston travel sensors. A communication protocol based on a subset of CANopen Draft Standards DS301 and DS401 has been developed to provide a standard architecture for integration of sensors and actuators into the OBMCS."""

Development and verification of in-vehicle networks include multiple design layers. These layers include the logical layer represented by the software application, the associated data link layer, and the physical connection layer containing bus interfaces, wires, and termination. Verification of these systems in the early stages of the design process (before a physical network is available for testing) has become a critical need. As a result, the need to simulate these designs at all their levels of complexity has become critically important. In-vehicle networks can be simulated on many different abstraction levels using various model levels and modeling technologies. Early in the development process, analyses can be performed without having available any detailed models from the chip manufacturer or component supplier. Later in the development process, more accurate models can be integrated into the simulation process, including those provided by suppliers and chip manufacturers. This paper demonstrates a portion of the development and verification process of the physical layer of an in-vehicle CAN Bus at Volkswagen using the Saber simulation environment. This paper also demonstrates integrating portions of the logical layer into the simulation so both logical and physical layers can be simulated together.

"""In-system programming (ISP) of flash memory in an embedded microcontroller is the state-of- the-art method for updating program code and application parameters This paper gives some background information on why a standard method for reprogramming via CAN bus is desirable, and how CANopen provides the necessary framework for it. This paper also presents an existing implementation of a CANopen bootloader. The CANopen bootloader is compatible with CiA standard DSP-302 which means that any master node or an SDO client or CANopen configuration tool running on a PC can update the firmware of such a node, using SDO write access to an Object Dictionary domain entry."""

Current problems with the interface between the truck and the superstructure will be explained and the requirements for an “ideal interface” will be defined. The requirements with respect to flexibility and real-time capability recommend the use of a field bus system. Reasons for choosing CANopen protocol will be explained. This will be followed by information about the structuring of the CANopen Truck Gateway standard, explaining also default communication and safety concept. In addition, the integration of the interface into the entire system of truck and trailer will be highlighted as well as special objects and mechanisms for superstructure manufacturers. Finally current activities with application specific associations are mentioned. Their focus is to ensure an application specific interface setup in an plug & play manner offering significant rationalisation.

Time-triggered CAN (TTCAN) combines the advantages of event- and time-triggered communication in order to fulfil the requirements of distributed real-time systems. Of crucial importance is thereby the generation of the communication schedule which should consider the demands of the time-triggered system on the one hand, while maintaining a good real-time performance for the event-triggered part of the system on the other. This paper deals with heuristic scheduling concepts for TTCAN networks and carries out a comparison by means of the above mentioned criteria. The suitability of the concepts is evaluated by laboratory measurements. The results enable the derivation of important clues in order to schedule TTCAN networks.

The media problem is one of the major causes of DeviceNet node failure in factory floor application; and one survey in a typical industrial application shows that a great percentage of DeviceNet system failure is caused by media problem. This paper presents an approach of designing DeviceNet media redundancy; this approach will improve the reliability of DeviceNet system; meanwhile, it is a simple and low cost solution that does not change current DeviceNet hardware and firmware.

The method of signal propagation analysis in the transmission lines of CAN-buses has been presented in the paper. Its aim is to determine and/or verify of the CAN controllers configuration parameters from the point of view of realistic conditions in the designed network which are often omitted in the standard calculation procedure. The adequate computer programs, to solve the transmission line equations in the area of assumptions, were elaborated. It allows to obtain the practically useful parameters like: characteristic line impedance, signal propagation times, voltage drops and ground potentials, etc, which, in turn, are selection base of right CAN controller sets. This paper contains the review of applied assumptions and examples resulted from the conducted investigations.

"""This paper proposes extending the high performance Network Processor (NP) architecture to meet cost/performance requirements of new emerging high speed and higher cost networks for the automation and automotive industries. The use of NP architectures for implementation of IDB-1394, MOST and Power Ethernet networks, which commonly embed CAN bus control elements, is proposed. The paper describes the implementation of a CAN enabled NP, incorporating a CAN co-processor within an Intel IXP NP hardware emulation system, and supporting the CANopen protocol. The proposed NP would enable packet transfer between CAN-bus and a range of existing NP communication protocols, including Ethernet, HSS, USB, PCI and UART."""

This paper presents a novel architecture and verification model of the CAN protocol controller for ASIC implementation. The key features of the proposed CAN controller are flexibility in terms of interfacing with host processors and smaller chip size. Also, the architecture is efficient for Intellectual Property (IP) reuse because of its flexibility and synthesis efficiency. For verification of the designed controller, we developed a verification model for fast verification during the design phase. The gate counts of the core logics in the proposed CAN controller are 3189 gates which are much smaller than other controllers. After successful verification, the CAN controller was fabricated by using 0.35_CMOS process.

CAN system designers are being challenged to meet stringent Electromagnetic Interference (EMI) and Electrostatic Discharge (ESD) standards while increasing reliability and reducing the size and cost of their products. To solve this dilemma, noise reduction techniques and bus protection devices can be implemented without significantly adding to the cost and complexity of the transceiver circuit.

"""This paper presents a cost effective CANopen implementation of a system to be de- ployed in an extreme environment exceeding MIL-STD operating temperature range of -55C to 125C. Applications such as Deep Oil Field exploration tools require robust electronic systems that can operate at 200C and beyond which very often results in the miniaturization of the electronics. To operate in this severe and noisy environ- ment, a CANopen network for inter-module communication was chosen. Contrary to a standard implementation, a CPU-less approach was selected where the entire CANopen stack is implemented in a Field Programmable Gate Array (FPGA). The FPGA is a System-on-Chip (SOC) integration where the CANopen stack is coupled with a CAN controller and peripheral logic for analog and digital I/Os. This CPU-less approach meets the size, cost, reliability, and power consumption requirements."""

"""Developing and testing of distributed embedded real-time control- systems is known to be very challenging due to the difficulties of de- bugging these systems in a target environment (e.g. due to weak moni- toring capabilities and lack of powerful debugging tools). The simulation technology described in this industrial experience paper is a toolbox aimed to improve the development and testing of distributed, CAN-based, embedded real-time control-systems. When using our technology, a complete control-system can be developed and tested without, or with only partial, access to target hardware. This is achieved by replacing target hardware dependent operations (e.g. device driver and operating system calls) with simulated equivalences that allow execution in a regular PC environment using regular PC pro- gramming tools. Thus, powerful PC tools for debugging, automated testing, fault injections, and dynamic modelling of the target machine, are made available for the embedded systems engineer. Complex dy- namic behaviours can be studied in the simulated environment, with- out access to the target hardware, e.g. allowing single stepping through scenarios. Simulating the complete system also facilitates customer tests and end-user evaluation of the system in an early phase of system devel- opment. It also shortens the turn around time for change, test, and evaluation, because development can be performed on a single PC in- stead of a full target system."""

The bit-stuffing mechanism utilised by CAN causes the message transmission time to become (in part) a complex function of the contents of the data fields. This variation in transmission times makes it difficult to predict the precise behaviour of real-time systems implemented using CAN. Previous work in our laboratory has led to the development of a software-based compensation method which significantly reduces the impact of CAN bit stuffing on message transmission times. In the present paper, we focus on the impact of bit stuffing on a system implemented using a “Shared- Clock” scheduling method. We use a detailed Hardware-in-the-Loop (HIL) testbed to explore the behaviour of an Adaptive Cruise Control (ACC) system for use in a passenger car. Through the use of the testbed, we present quantitative results which demonstrate the impact of variations in the message transmission times on the performance of the ACC system. We go on to demonstrate the improvement in performance which results when the previously-mentioned compensation technique is employed. Finally, the memory and CPU resources required to implement this compensation are discussed.

The CLAN intellectual property core is a CAN 2.0B controller developed at the Electronics and Telecommunications Department of the University of Aveiro, for research and educational purposes and in particular with the aim of providing the adequate hardware support to implement and validate higher layer protocols such as TTCAN or FTT-CAN. It was modeled at RTL level using the VHDL hardware description language, synthesized, implemented and tested on Xilinx FPGAs. However, the model is technology independent and can be synthesized for different implementation technologies from FPGAs to ASICs. The CLAN IP core fully implements the CAN 2.0B specification and it includes also a synchronous parallel microprocessor interface, interrupt generation logic and some advanced features, such as message filtering, single shot transmission and extended error logs and statistics. The data bus width can be 8, 16 or 32 bits wide. For applications where microprocessor interface is not needed or a different interface is required, the core internal module that implements the protocol can be used separately. The CLAN controller with microprocessor interface logic occupies about 30% of a Xilinx Spartan-IIE XC2S300E FPGA, corresponding to 100,000 equivalent logic gates, approximately. It was tested with other commercial controllers within a bus operating at 1Mbit/seg.

"""One of the challenges in today’s communication applications is that more and more different network technologies need to be interfaced to another. A machine might use an RS-485 or other general serial network and also requires access to a higher control level via networks such as CANopen. The interfacing between those networks is especially challenging if not only different communication technologies but also completely different network protocols are used. Gateways interfacing such different networks need to process data through all protocol layers – in each direction in and out of the gateway – making this a higher-end application in terms of MCU performance, memory requirements and software development. Interfacing between networks would greatly be simplified if the same network protocol would be used on the various communication technologies as illustrated by figure 1. Instead of complex gateways, simpler bridges could be used requiring far less resources and development work. ?One of the few network protocols truly open to multiple communication technologies and applications is CANopen. It was originally developed to operate on CAN, however it was always kept open enough to be used on multiple communication technologies and with any application. This paper and class examines the core features of CANopen and how they can be adopted to other communication technologies"""

"""This paper details the first implementation of MilCAN on a UK military vehicle. The UK Challenger 2 (CR2) Main Battle Tank manufactured by BAE Systems Land Systems has been updated to include improved Commander’s Situational Awareness (SA) with new technologies communicating over MilCAN. Situational Awareness is critical to any fighting force allowing Battlefield Commanders to make better tactical decisions in a reduced timeframe, whilst providing the Commander with mission planning capabilities such as generation and monitoring of orders, reports, overlays etc. Currently, command and control on the battlefield is based largely on manual processes for the monitoring and planning of operations. The introduction of the Platform Battlefield Information System Application (PBISA) System for CR2 interfaces with both new and legacy vehicle sub-systems and allows automation of many of the currently manual tasks. The MilCAN data bus provides the communication between several processing resources on the vehicle and presents the vehicle Commander with essential information. The MilCAN data bus also distributes accurate and up to date position data to both Commander and Driver at their respective displays in a simple and easy to read format."""

"""The Controller Area Network (CAN) has seen enormous success in automotive body and powertrain control systems, and in industrial automation systems using higher layer protocols such as DeviceNet and CANopen. Now, the CAN standard ISO11898 are being extended to Time Triggered CAN (TTCAN) to address the safety critical needs of first generation drive-by-wire systems. However, their successful development depends upon the availability of silicon and software support, and appropriate development & analysis tools. This paper outlines some of the exciting developments at Warwick Control Technologies which includes:- a) An implementation that can run deterministically at 60% bus loading at 1Mbit/s and 70% at 500Kbit/s (almost twice that of traditional automotive CAN systems), b) Interoperability testing with the Bosch c) Rapid control prototyping for a TTCAN Embedded Node, and d) TTCAN transmitter/fault injector tool."""

"""During the current development of a modular drivers desk for trains a major factor that has to be considered is the input and wiring of the control signals. There are up to about 45 controls situated over the complete desk. The CAN (open) Bus system was selected for this purpose in order to maintain a flexible approach with a minimum of costs . When considering railway applications it is very important that all the standards and specifications (EN50155 etc.) are complied with and that the system guarantees an extremely competent method of control signal transfer. The controls must provide a high availability so that failures very rarely cause a break down. The paper describes the principles behind a 2 channel CAN System with control input backup via the display terminals of the dashboard in the cab. The aim is to maintain control over the main important input functions despite failures."""

In every day R&D work employing network technologies like e.g. CAN proper shielding often is impossible as frequent adjustments and new configurations have to be applied to the systems and therefore the outer system surface has to be kept open. Therefore EMI problems are commonplace. This leads to the problem that a lot of effort is distracted from the development of the circuit itself and spent on dealing with the resulting CAN errors. A very cost efficient way to avoid such problems is to use a low-cost optical CAN. For this only the physical layer has to be changed and these changes are such that existing electrical CAN segments can be merged with the optical segments. In such a configuration the hub or switch which is needed for these star topologies must be suited for real-time error detection in optical segments to avoid the breakdown of the whole network. Also the hub or switch should be able to support real-time connection or disconnection of the node when CAN nodes are connected or disconnected to the network, to avoid bus conflicts with the other network sharing units. We have developed a low-cost optical CAN network and have successfully implemented it in a hybrid drivetrain testbed. The previous problems of destroyed CAN messages due to heavy EMI resulting from power electronics like DC/DC converters were solved. Electric network segments were merged with optical segments.

"""CANopen is a well proven, well established and very versatile Fieldbus technology, implemented in a large variety of devices. CAN unquestionably has distinct advan- tages just as low connection costs per device, true multi-master capability and out- standing error detection and handling features. However, for demanding applications like motion control or applications which require large network extension CAN is in- creasingly challenged by the upcoming industrial Ethernet technologies. It is shown that there is an Ethernet technology that retains CANopen to such a large extend, that even most of the CANopen communication protocol stack can be re-used. CANopen over EtherCAT features PDOs and SDOs as well as an Object Dictionary, the CANopen state machine and fully supports all CANopen device profiles. It is shown that EtherCAT provides a smooth migration path for CANopen devices towards the brave new world of Industrial Ethernet."""

"""Single-bus CANopen systems have not been enough for large systems, where CANopen allows use of existing tools and good selection of standardized components. Still CANopen documentation is single-bus oriented, but does not limit the systems into single bus level. Many things, like signal oriented communications model, support more complex systems. Typically there are more than enough maximum supported amount of nodes and COB-Ids but not enough transmission bandwidth to carry all signals of the system in the desired time windows. This paper presents one solution, where every bus is still 100% CANopen conformant. Dividing a large system into smaller subsystems makes it easier to schedule signals into PDOs, to monitor buses by keeping subsystem specific signals local to appropriate subsystem and to decrease the coverage of fatal bus errors. Special gateways are introduced for signals, SDO- and EMCY-protocol. Also protocols and services restricted into one hierarchy level and reasons for the restriction are explained. The biggest challenges are found in EMCY-protocol and signal scheduling over multiple independent bus segments. In those issues there are also the biggest risk for application dependencies. Hopefully this paper is a trigger for official documentation and advanced development of multilevel CANopen networking."""

"""The routing of messages across CIPTM1 networks (CIP = Common Industrial Protocol) is one of the most interesting features of CIP since it allows seamless transition of messages (connected and unconnected) between a DeviceNetTM2 network and other networks of the CIP family, a feature that is not found in any other industrial communication networks. This contribution explains the following details of the mechanics of the routing process for both connected and unconnected messages: • Explanation of the general principle of the routing process • The port object and port segments • Details of unconnected message routing • Details of connected message routing • Execution of the routing • Object addressing and object visibility within routers as seen from multiple networks • Route browsing through multiple networks • Details of the routing within DeviceNet • Route representation with Electronic Data Sheets (EDS) • Bridging into non-CIP networks (last hop) • Real-world example: Full details of a trace of explicit messaging The advantage of the described method is the fact that there is only one routing method within all CIP networks (DeviceNet, ControlNetTM3, EtherNet/IPTM4) and therefore, only minimal translation of messages is required."""

"""A UML based framework for the simulation of distributed systems of embedded control units (ECUs) has been developed by proTime. This framework generates at its runtime an executable model of a system, described by an editable system description. Thereby this simulated system can be used for tests and analysis. A basis for this UML based framework are executable models of different field busses. Currently models exists for CAN and FlexRay, Wakeup Line and a real-time control bus for simulation control purposes. These field bus models will be used (even simultaneously) to connect the modeled ECUs while the modeled system is under simulation. This UML based framework allows both the simulation of communication processes, and the simulation of functionality depending on these communication processes all with the correct timing. The main advantage of this UML based framework is the testability of communication dependant functionality like gateways, control loops and network management without an expensive test environment. The principal theme of this speech will be the CAN bus model, its integration into the runtime-created system model and the resulting (almost boundless) possibilities. Currently automotive systems are the focus of our modeling and simulation, but the applicability of this UML based framework is not restricted to these tasks."""

"""Bus architectures with up to five independent CAN channels are used in today's auto- motive and industrial control systems. Caused by the rising numbers of sensors, actu- ators and electronic control units over the last years, modern control concepts demand devices supporting cross-linking of these channels. This interconnection is realized with a CAN gateway that connects several CAN buses between sub networks at differ- ent speeds. Current gateway implementations are based on one of two concepts. The one concept is an application-specific multi-channel CAN controller with shared message object memory. This concept is inflexible regarding the gateway structure, especially the number of CAN channels, but it enables the transfer of messages between the net- works without causing a high load on the host CPU. The other concept is a set of single channel CAN controllers served by a message handling software on the host CPU. This implementation is more flexible regarding the gateway structure, but the load on the CPU depends on the combined bus traffic of all connected CAN networks. Starting from these two solutions, a new concept has been developed, combining the advan- tages of a flexible structure with a low CPU load. In this paper, the three concepts are compared and advantages/disadvantages are shown. In addition, problems in the design of gateways are discussed."""

"""Electronic network-centric devices are being widely used in military vehicles. The environments they operate are demanding and performance delivery is crucial under harsh conditions. Dictated by application needs, customised levels of configurability control and analysis are essential. The objectives of the work presented in this paper are to provide intelligent means of connectivity for integrating diverse network technologies, and tools for configuring and monitoring military vehicle electronic systems. Our investigation introduces the use of smart bridges that allow access to custom tools for reconfiguration and network information from the system. The target is to enhance the means of configuring these bridges, and to acquire customised diagnostics that potentially provide enhanced performance solutions. Experimental investigations utilizing MilCAN and Ethernet are being conducted, using a variety of different network topologies and configuration sets, to interconnect multiple MilCAN segments over an Ethernet backbone."""

In the past, CANopen networks were used mainly as stand-alone networks embedded in machines and deeply embedded in sub-systems. If required, non-standardized gateway devices realized the connection to higher-level Ethernet-based networks. In future applications, Ethernet-based networks also may be connected to higher-level CANopen integration networks. Besides hierarchical network architectures, there may be required non-hierarchical network architectures, too. The paper presents the cur- rent status of standardized CANopen gateways to Ethernet-based networks, and dis- cusses the requirements on further standardization activities.