Professional Documents
Culture Documents
Replacement of The Controller Area Network (CAN) Protocol For Future Automotive Bus System Solutions by Substitution Via Optical Networks
Replacement of The Controller Area Network (CAN) Protocol For Future Automotive Bus System Solutions by Substitution Via Optical Networks
Replacement of The Controller Area Network (CAN) Protocol For Future Automotive Bus System Solutions by Substitution Via Optical Networks
1. INTRODUCTION
In the near future it is likely that vehicles will operate without a driver or conductor. Vehicles will be able to
transport individuals autonomously to a desired destination with just a few initial inputs. As good as that may
sound, the technical effort to realize such a system with minimal error margin, is immense.
In general, vehicles are growing increasingly sophisticated. For example, Advanced Driver Assistance Systems
(ADAS) like Traffic Sign Recognition, Lane Keeping Assist System (LKAS), Collision Avoidance System,
Automatic Parking and even Vehicular Communication Systems are currently or will be available in state-of-the-
art vehicles. However, with greater sophistication comes greater complexity of those applications, systems and
networks. Regardless of the basic requirements, such as real-time, safety, security etc., certain conditions have to
be met. Otherwise, it is impossible to ensure the functionality of an application, the vehicle, or in a wider sense,
even the physical integrity of the passenger. Moreover, a person must all times be able to take over control of the
vehicle at any point to fulfil valid regulations.
One example for the growing complexity is the amount of wiring that increases with every single addition of
an (assistance) system to the vehicle. Every sensor, actuator and camera must be connected to a central control
unit. Those additions increase the weight of the vehicle, decrease the available space and bring more elements
into the vehicle, which might cause unforeseen interferences. Such interferences, mainly of an electromagnetic
nature, are ubiquitous in a vehicle, where mostly electrical components are installed. This is not only true for
automotive applications but also even more in airplanes or space-faring vehicles where any kind of interference
can pose a serious threat to the vehicle and its crew. Therefore, those influences must be taken into account,
especially in the design phase.
In this paper, we will consider existing bus systems as well as their glitches and challenges. Based on this
analysis, an approach for a novel optical network will be presented.
system; the system is not synchronized by a trigger or time. The general idea was to use one single wire for
several connections to reduce the amount of wiring to a minimum [1], hence save space and weight.
For a simple depiction of a CAN bus see Fig. 1.
device A device B device C
bus
termination impedance termination impedance
device D device X
2
ICTON 2016 Mo.D6.1
3
ICTON 2016 Mo.D6.1
The low transmission speed of CAN is limiting for future application and systems. With a data rate of
125 kbit/s (CAN-Low-Speed-Physical-Layer) up to a maximum of 1 Mbit/s (CAN-High-Speed-Physical-Layer),
only the most important data can be transmitted. Hence, those systems are limited to a certain amount of control
units, which can be part of such a bus – considering the increase of elements like sensors and other actuators.
Yet another problem of CAN are the short message lengths, which limit the transmission to a maximum of
130 bit (standard format). With this message length, the transmission data rate of up to 1 Mbit/s is sufficient.
However, if longer messages without any collisions or other issues should be sent, the data transmission rate is
not adequate. If traffic on the bus is increasing, the data transmission will take considerably longer and the bus
system will not fulfil the defined, and therefore crucial, requirements anymore. Since automated driving is the
targeted goal, it is also important to provide the system with real-time capability. In dangerous situations, every
ms or even ns could count to prevent a potential accident. Compared to the reaction time of a human driver,
automated/autonomous vehicles will always achieve better results.
Current CAN busses are not working in real-time since they have a time delay of 1ms and up to 10ms. If the
traffic on the bus is higher, the delay will increase proportionally with the amount of data on the bus. More
sophisticated applications are limited by the CAN bus and cannot function in real-time. One example is the
advanced driver assistance system Lane Keeping Assistance, for which real-time and safety, as well as a high
data transmission rate, is of utmost importance [19]. The CAN protocol can only handle minor traffic without
collisions and is therefore unfeasible for the described future automotive system with such automotive
applications.
There is also the problem when assigning priorities to the messages, which are transported over the bus.
In most cases, no issue will occur for those messages with a low identifier. Messages with a low identifier will
most certainly be transmitted. Messages with a high identifier however might be delayed or suppressed
completely [20].
Considering all these problems, it is obvious to search for new solutions to cope with those challenges in the
future rather sooner than later.
4
ICTON 2016 Mo.D6.1
However, when using only a single fibre it is necessary to use different wavelengths. Otherwise, the light
waves might cancel themselves out or interfere with each other so that no communication is possible. The
distances within a vehicle are quite short so the use of multi-mode optical fibres should be sufficient.
For optical communications, several transmission windows have been defined where the attenuation of the
signal is the weakest. There are currently six different bands defined in the wavelength range of 1260 nm to
1675 nm [24]. It is possible to distinguish between different light waves when using a gap of 2 nm. Hence,
sufficient unique wavelengths for data transmission would be available for a complex system.
The distances that have to be covered with the fibres are in the range of several meters, therefore the
attenuation is considerably small. Considering all influences on the bus, including a wide range of temperatures
[-40°C to +85°C], the worst-case fibre attenuation is approximately 0.4 dB/m [12] and confirms that optical
busses are very useful for automotive applications.
4.2 Basic Concept Using Optical Data Communication
Based on optical communications, a new concept for an automotive bus shall be presented. Optical transmission
technologies provide protection against electromagnetic inferences, which are and will be even more of
a problem in the future. The technology is promising to be a solution for the future data transmission challenges
that will appear in automated and autonomous vehicle systems. Since the amount of transmitters is increasing in-
and outside the vehicles, the chance of that interfering signals occur is as well.
It is possible to reduce the amount of cabling to a single optic wire with several fibres inside to transport the
information. In this case, it is required to use wavelength multiplexing, to prevent any reflections on the wire.
To quantify the weight savings, the information provided by car manufacturers is considered. Copper cabling
in modern vehicles amounts to a total weight of 60 – 100 kg for almost 3 km [18, 25]. In comparison, the weight
for a commonly used optical fibre consisting of silica with less than 24 fibres amounts to 11 kg per kilometre
[26]. Thus, weight savings of at least 50% can be achieved. Even more so (approximately 90%), when using
Polymer optical fibres instead of drawing glass (silica), which are also considerably cheaper. Moreover, they
provide a major advantage in automotive applications since the fibres made out of polymer are more robust when
stretched or bent. However, the attenuation and distortion are far higher and therefore it is only feasible for short
distances (less than 200 m) [27].
When developing a new concept, all parameters have to be considered in the designing phase. To proper
support such an approach, an exact positioning of the communication fibres as well as of the used devices is very
important.
4.3 Components and Topology
Several different components will be required when changing from copper wires to optical fibres. The whole
concept is based on using a central processing unit (CPU) or central control unit (CCU) which manages and
monitors all connected devices. This CPU/CCU is connected with the data bus, consisting of several fibres.
A single optical wire with several fibres inside should be sufficient to cover the communication within the
vehicle. Even more so, the achieved bandwidth will be at least 100 times faster than with the commonly used
CAN bus.
The first step is to change the topology from a bus topology to a hybrid of bus and star. The reason for this
action is to ensure the flexibility of the network because optic fibres lack the potential for simple extensibility.
Therefore, a design decision must be made beforehand to ensure flexibility can still be achieved.
The structure of the network is shown in Fig. 2.
device sensor ... device sensor sensor sensor
Central Processing Unit
switch
optical bus
5
ICTON 2016 Mo.D6.1
Lock
Light
Instruments
Central body control
Sensors Sensors
6
ICTON 2016 Mo.D6.1
Datagram Protocol (UDP) and TCP – message-orientation for reliability and in-sequence transport of messages
with congestion control [30].
Problems of TCP stated in the specifications of SCTP [30]:
− TCP provides both reliable data transfer and strict order-of-transmission delivery of data. Some
applications need reliable transfer without sequence maintenance, while others would be satisfied with
partial ordering of the data. In both of these cases the head-of-line blocking offered by TCP causes
unnecessary delay.
− The stream-oriented nature of TCP is often an inconvenience. Applications must add their own record
marking to delineate their messages, and must make explicit use of the push facility to ensure that
a complete message is transferred in a reasonable time.
− The limited scope of TCP sockets complicates the task of providing highly-available data transfer
capability using multi-homed hosts.
− TCP is relatively vulnerable to denial of service attacks, such as SYN attacks.
In a vehicle, the order of transmission is not of importance. However, it is crucial that certain messages are
transmitted faster (with a higher priority) than others. This is possible with SCTP since this protocol assigns
a sequence number to each message sent in a data stream. TCP assigns only one sequence number to a whole
segment, so no real distinction between single messages can be made. Ordering messages is optional in SCTP, but
proves very useful for automotive applications since assigning priorities is required for accurate functionality.
Another benefit of SCTP over TCP is that it provides redundant paths to increase reliability [30]. SCTP uses
heartbeats to check if a connection is still valid. If one node fails, the connection will find another way,
if available. The SCTP protocol has some additional security features and adaptability, which might come in
handy for Car-2-X communication in the future.
REFERENCES
[1] M. Farsi, K. Ratcliff, and M. Barbosa: An overview of controller area network, Comp. and Control Eng. J.,
vol. 10, no. 3, pp. 113-120, Aug. 1999.
[2] The Differences Between Fluid Control and Electric Actuators, Feb. 11, 2016 by M. Hyche
http://empoweringvalves.com/the-differences-between-fluid-control-and-electric-actuators/, last accessed
date 12/04/16.
[3] CiA – CAN in Automation: http://www.can-cia.org/, last accessed date 10/04/16.
[4] CAN Specification 2.0, Part A, https://web.archive.org/web/20141121173756/http://www.can-
cia.org/fileadmin/cia/specifications/CAN20A.pdf (substituted by ISO 11898-1), last accessed date
12/04/16.
[5] F. Hartwich, B. Müller, T. Führer, and R. Hugel: CAN Network with Time Triggered Communication,
Robert Bosch GmbH,
http://www.datamicro.ru/download/iCC_07_CANNetwork_with_Time_Trig_Communication.pdf, last
accessed date 11/04/16.
[6] F. Hartwich: CAN with Flexible Data-Rate, Robert Bosch GmbH, http://www.bosch-
semiconductors.de/media/ubk_semiconductors/pdf_1/ipmodules_1/can_fd/icc13_2012_paper_Hartwich.pdf,
last accessed date 10/04/16.
[7] E. Mayer: Serielle Bussysteme im Automobil, Flex Ray für den Datenaustausch in sicherheitskritischen
Anwendungen, Elektronik automotive 4.2012, pp.13-16,
http://vector.com/portal/medien/cmc/press/PressReport-SerielleBussysteme-DE.pdf, last accessed date
12/04/16.
7
ICTON 2016 Mo.D6.1