WO2016038225A1

WO2016038225A1 - Method for an improved throughput of sensor data in a communication system, and corresponding communication system

Info

Publication number: WO2016038225A1
Application number: PCT/EP2015/070986
Authority: WO
Inventors: Aleksandar Koenig; Thomas W. Kelly; Richard Baur
Original assignee: Analog Devices Global; Bayerische Motoren Werke Aktiengesellschaft
Priority date: 2014-09-12
Filing date: 2015-09-14
Publication date: 2016-03-17
Also published as: DE102014113162B4; DE102014113162A1

Abstract

The invention relates to a method for transmitting data in a communication system (100), having the steps of (a) sending a first synchronization signal (411, 511, 611, 711) from a bus controller (150) to at least one sensor controller (110), (b) detecting an analogue signal by means of a sensor (111) of the sensor controller (110), (c) capturing at least one signal value of the detected analogue signal at a predefined instant after the first synchronization signal (411, 511, 611, 711), as a reaction to the first synchronization signal (411, 511, 611, 711), (d) converting the at least one signal value into a sequence of binary signals, (e) filtering the sequence of binary signals by means of a low-pass filter (113) having a predefined cut-off frequency in order to produce a digital value, (f) transmitting the digital value from the buffer store (114) of the sensor controller (110) on a communication system, particularly on a communication bus, wherein steps (b) to (g) are performed N > 1 times, as a reaction to step (a).

Description

METHOD FOR AN IMPROVED THROUGHPUT OF SENSOR DATA IN A COMMUNICATION SYSTEM, AND CORRESPONDING COMMUNICATION

SYSTEM

The present invention relates to the transmission of data in a communication system, particularly in a synchronous real-time communication system.

Communication systems are known in the prior art. They are used to transmit data from one subscriber in a communication system to another subscriber. The sum total of the rules for the data transmission is called a protocol .

Communication systems are normally distinguished according to synchronous and asynchronous communication systems. In this context, a synchronous communication system has all the actions of the subscribers related to a central clock, and in the case of asynchronous communication systems the actions of the subscribers are triggered by one or more predefined signals from a subscriber .

Real-time communication systems have a series of rules that guarantee an upper limit for the response time for each subscriber. This guaranteed response time means that - besides other restrictions - the frequency with which each subscriber can transmit data is limited by the protocol. By way of example, in many real-time communication systems, each subscriber can send only once within a time period T_p.

Such real-time communication systems according to the prior art have the following disadvantages, inter alia: if relevant data accumulate more frequently than at a frequency 1/T_P, they can no longer be transmitted as often as necessary without the protocol of the communication system needing to be changed. However, a change in the protocol would mean that it is no longer possible to use standard components - operating according to the standard protocol. This would require replacement of the whole system, which would firstly entail high costs and would secondly be inconsistent with the approach of evolution in small steps that is not uncommonly chosen in safety- relevant realms.

Against this background, it is an object of the invention to provide a communication system that at least partly overcomes or improves the disadvantages of the prior art that are outlined above and, in particular, allows the data throughput in a communication system to be increased, preferably without this requiring the protocol to be changed.

This and further objects are achieved by a method according to independent Claim 1 and a communication system according to independent Claim 15. Embodiments and developments of the invention are the subject matter of the subclaims.

The method according to the invention is used to improve the data throughput for the transmission of data in a communication system, particularly in a synchronous communication system, preferably without changing the communication protocol, so that the use of standard components continues to be possible. To this end, the method according to the invention involves a first synchronization signal being sent from a bus controller to at least one sensor controller in a first step (step a) ) , as in the case of the standard protocol. In one embodiment, the synchronization signal is sent to a plurality of sensor controllers. In a second step (step b) ) , an analogue signal is sampled by means of a sensor. The time profile of the analogue signal does not in any way need to be correlated with the synchronization signal from the bus controller in this case.

At a predefined instant after the first synchronization signal is sent, at least one signal value of the sampled analogue signal is captured in a further step (step c) ) of the method according to the invention. By way of example, the capture can be effected by means of a sigma/delta modulator, with the analogue signal being converted into a plurality of binary signals, and particularly a sequence of binary signals, in a subsequent step (step d) ) .

In the next step (step f) ) of the method according to the invention, the digital value is transmitted from the buffer store of the sensor controller on the communication system, e.g. to any other subscriber or to a central entity that - on the basis of the transmitted values from one or from a plurality of subscribers - performs further processing. In one embodiment, this communication system may be a synchronous communication bus in which the method according to the invention can advantageously be applied.

According to the invention, the sensor controller performs steps b) to f) N > 1 times, as a reaction to step a) .

Usually, the plurality or sequence of binary signals that is produced by a sigma/delta modulator is a one-bit signal with a very high sampling frequency ( "oversampling" ) . Such signals are normally not transmitted directly via the communication system. Rather, the signals are filtered in a subsequent step (step e) ) by means of a digital low-pass filter and integrated over a particular period of time that is correlated with the cut-off frequency of the low-pass filter.

The predefined instant for capture of the signal value or for the start of conversion, particularly by means of a sigma/delta modulator, can be defined in the method according to the invention by the bus controller or the sensor controller in the form of software or hardware in an initialization phase, for example. In this case, the predefined instant is defined by the bus controller as a reaction to the synchronization signal, and normally at a time relative to this synchronization signal, wherein the interval of time from the synchronization signal may also be essentially zero, so that capture of the signal value and the processing that follows that thus take place essentially at the same time as reception of the synchronization signal by the sensor controller.

The predefined instant is usually defined differently for each of the N > 1 executions of steps b) to f) , as a reaction to step a) . The bus and/or sensor controller therefore have "knowledge" of the instant in relation to a synchronization signal at which at least step c) and optionally also steps d) to f) take place for each pass of the N > 1 executions of steps b) to f) , as a reaction to step a) .

The advantage of the method according to the invention is that the sensor controller can transmit the N > 1-fold number of digital values after reception of the synchronization signal in step a) . Such an increase in the transmission rate is useful or necessary particularly when the boundary definition of a limiting element is raised. The signal propagation time can be shortened in accordance with the object by increasing the cut-off frequency of the low-pass filter. In order to be able to use a low-pass bandwidth increased in this manner, the sampling rate can be increased.

In the method according to the invention, an increase in the cut-off frequency of the low-pass filter results in the N > 1-fold number of digital values being able to be transmitted so that the whole system can use the additional information. In this case, there is a close correlation between the number N of transmissions and the cut-off frequency of the low-pass filter. The synchronization signal is usually a relatively short pulse, the precise duration, voltage profile, tolerances, etc., of which are stipulated by the standard protocol. For the method according to the invention, all actions that take place after step a) are related to step a) , specifically both in terms of logical and chronological sequence ("timing").

In one embodiment, the synchronization signal is sent repeatedly, as a result of which the method steps according to the invention are also executed repeatedly.

In particular, the synchronization signal can be sent periodically, with a period T_p. This is particularly advantageous when the sensor values are intended to be read and transmitted periodically, particularly in real time.

A bus controller or ECU (Electronic Control Unit) within the context of the invention is a central controller that controls the behaviour of all subscribers in the communication system, by means of the sensor controller. A sensor controller within the context of the invention is a controller that every single subscriber in the communication system has. The sensor controller supports the protocol of the communication system and thus forms the counterpart of the central bus controller. In addition, it controls the sensor and the conversion of the analogue sensor values into a digital value.

In an alternative embodiment of the method according to the invention, the signal value is captured by means of a sample-and-hold apparatus that "freezes" the signal value at the predefined instant in order to produce a digital value therefrom by means of an analogue-to-digital converter (A/D converter, ADC) . In this case, the (analogue) low-pass filter is applied before the signal value is "frozen"; the filtering is thus performed before step c) .

In one embodiment of a method according to the invention, the plurality or sequence of binary signals and/or the digital value are stored in a buffer store of the sensor controller (or buffer) . In this case, the buffer-storage can be applied both to the plurality or sequence of binary signals - before they pass through the low-pass filter - and to the digital value obtained, before said digital value is transmitted on the communication system. This results in decoupling of the timing of the signal sampling and conversion of the transmission of the data on the communication system.

In one embodiment of a method according to the invention, steps b) to f) are performed N > 1 times, as a reaction to step a) , before a second synchronization signal is sent .

This embodiment has the advantage that the communication system can therefore be of particularly simple design. In particular, it provides a simple way of implementing a method according to the invention in a manner compatible with a standard protocol.

It can be expressly pointed out that an alternative embodiment of a communication system according to the invention can allow steps b) to f) to be performed not yet N times - but N > 1 times - before a second synchronization signal is sent. By way of example, a first and second transmission can take place before the second synchronization signal, but a third and fourth transmission can take place only after the second synchronization signal. Such communication systems are called "pipelined bus", for example.

In one embodiment, the method according to the invention is implemented in a synchronous bus system. In particular, a multiplicity of protocols, e.g. both synchronous and asynchronous protocols, may be provided in the communication system in which the method according to the invention is implemented.

Particularly preferably, the synchronous bus can in this case allow the transmission authorization to "circulate" between the subscribers ("round robin"), i.e. with M possible subscribers - on M sensor controllers - each subscriber 1 to M can send one message each after reception of the synchronization signal. For this, one time range ("timeslot") Tl to TM each is reserved for each subscriber. If the period of time between two synchronization signals is T_p and all the subscribers send one message each before a second synchronization signal is sent, then each subscriber has a timeslot of maximum duration T_p/M. In this case, the i-th subscriber always has the i-th timeslot assigned to it. The number of actual bus subscribers m can be no more than the number of possible subscribers M (m < M) . If it holds that m < M, then at least one timeslot is free in the prior art. Furthermore, in the method according to the invention, the i-th subscriber can, in principle, send not only in the i-th timeslot but also in another timeslot, for example in the j-th and/or k-th, etc. timeslot . In one embodiment, the method according to the invention is implemented in a communication system according to the Peripheral Sensor Interface 5 (PSI5) standard. This bus standard is regularly used in sensor systems, particularly preferably in sensor systems for vehicles, particularly for cars.

In this embodiment of the method according to the invention, it is particularly important that the implementation cooperates with bus controllers and sensor controllers of any PSI5 communication system according to the prior art without difficulty. Nevertheless, there is a need to increase the transmission bandwidth on the communication system at least for some sensors, which can be accomplished by the method according to the invention without this infringing the PSI5 specification.

In a further embodiment of the method according to the invention, the low-pass filter has a cut-off frequency that correlates with the number N of transmissions according to step f) , preferably correlates linearly, and particularly is approximately 800 Hz or approximately 1600 Hz. The increase in the cut-off frequency of the low-pass filter also results in a reduction in the group delay time of the filter. In one development of this embodiment, the low-pass filter is a second-order or higher-order low-pass filter. This allows a better signal-to-noise ratio to be achieved.

In this embodiment, the advantage according to the invention becomes particularly clear: if, by way of example, the communication system used is assumed to be a PSI5-based system, which is operated as a synchronous bus according to sections 2.4 and 6.8 of the "PSI5 Technical Specification", then the interval between two synchronization signals is T_p = 500 \ls . This results in a maximum sampling rate of 1/500 \ls = 2 kHz. According to the sampling theorem, a maximum cut-off frequency of 1 kHz is therefore possible.

In the case of some sensors - for example in the case of crash sensors - strong signals (e.g. factor 100 or 1000 in comparison with an ordinary signal level) can arise in a frequency range above the cut-off frequency of the low- pass filter, as a result of which the low-pass filter does not fall steeply enough to filter such signals before sampling. Therefore, for one class of sensors, a lower cut-off frequency is chosen for the low-pass filter than is necessary when exclusively taking the sampling theorem into consideration. By way of example, when taking said effects into consideration, a cut-off frequency of 400 Hz can be chosen for a maximum sampling rate of 2 kHz .

At this juncture, it should be noted that these advantages may also be advantageous for communication systems other than PSI-based systems, and therefore the method according to the invention can also be used therein .

In particular, the cut-off frequency of the low-pass filter in a communication system implementing the method according to the invention and having periodically repeating synchronization signals with a period T_p can be chosen, according to the invention, such that said cut^¬ off frequency is greater than 1/ (2*T_P) , preferably is equal to a/ (2*T_P) , where a > 1. At the same time, a can correspond to the number of transmissions N within the period of time T_p.

For a particular class of sensors, it is possible - for the reasons explained above - to choose a lower cut-off frequency than would be necessary when exclusively taking the sampling theorem into consideration. By way of example, when the aforementioned effects are taken into consideration, a cut-off frequency for the low-pass filter of 400 Hz can be chosen for a maximum sampling rate of 2 kHz or a cut-off frequency for the low-pass filter of 800 Hz can be chosen for a maximum sampling rate of 4 kHz .

For some sensors, better values are advantageous, e.g. it is advantageous to transmit the values of the acceleration sensors of cars more frequently and more quickly in order to optimize the trigger times of the airbags in the event of an accident and therefore to increase the safety of the vehicle occupants. If the increase in the cut-off frequency of the low-pass filter means that these values can be transmitted more frequently than once between two synchronization signals in the period of time T_p, then the rise in the sensor values can be identified more precisely and at the same time this method can improve the reaction time in the event of an accident.

In a further embodiment of the method according to the invention, the digital value has a timestamp added to it, particularly a timestamp for the capture of the signal value according to step c) . This timestamp is then optionally transmitted together with the digital value in step f ) .

When a sensor controller transmits the sensor values N > 1 times according to the method of the invention, it is possible that strictly periodic sending of the sensor values is no longer guaranteed, i.e. the intervals between two transmission instants Ti and Tj are possibly no longer equidistant and/or the period of time between capture of the signal in step c) and sending in step f) is possibly no longer of equal length at transmission instants Ti and Tj. Thus, if the rise in a signal value between transmission instants Ti and Tj is intended to be stipulated, for example, this can lead to incorrect computation of the rise. To avoid this, the correct instant of capture can be restored by means of the timestamp .

In a further embodiment of the method according to the invention, an analogue signal is sampled a plurality of times, particularly more frequently than the sampled value is transmitted, by means of a sensor in the period of time T_p. These - to a certain extent "internal" sampling instants are known to the system. From the knowledge of the "internal" sampling instants, it is then possible (even without a timestamp) for the exact instant of detection of an analogue signal to be known to the whole system. In one embodiment of the method according to the invention, steps b) to e) are performed by means of a sigma/delta analogue-to-digital converter (sigma/delta ADC) .

It is therefore not just possible for the ADC itself to be produced particularly inexpensively, rather the digital - filter can also be integrated better, e.g. on the same chip, and therefore be produced in smaller form, less expensively and such that it is less susceptible to changes in the surroundings (e.g. the temperature) .

The method according to the invention can be applied for a large group of sensors: preferably, the sensor is selected from a group that comprises angle sensors, pressure sensors, acoustic sensors, optical sensors, magnetic sensors, particularly Hall sensors, speed sensors, current sensors, voltage sensors, capacitive sensors, resistive sensors, acceleration sensors, particularly crash sensors, preferably capacitive crash sensors .

The sensors may be implemented on a separate assembly, as a separate component on the same assembly or else on the same component, for example as a microelectromechanical system (MEMS) . In one embodiment, an optical sensor can transform temporal fluctuations in the brightness into the output values "change in a voltage over time"; an acceleration sensor can change its nonreactive resistance value by means of a strain gauge, for example, or, for example in a MEMS, can cause a change in the electrical capacitance on the basis of the acceleration. Hence, a broad field of application is possible for the method according to the invention, even if there is preferably provision in the present case for application in the context of the transmission of data in a real-time system, particularly in a vehicle, for transmitting the data from crash sensors.

In one embodiment of the method according to the invention, a single instance of the sensor controllers behaves like two sensor controllers ("dual channel device") or a plurality of sensor controllers ("multichannel device") on the communication system. In the case of a PSI5 bus, such a sensor controller uses only one physical line, but behaves in the manner of two or a plurality of "logical" sensor controllers.

This achieves a reduction in the wiring because the sensor may therefore be on the same physical assembly or even on the same chip as all of its sensor controllers, or else a single chip can control a plurality of sensors. Two or more sensor controllers having a plurality of "logical" interfaces to the communication system thus have only a single physical interface in this embodiment.

A communication system according to the invention for transmitting data has the following components:

• A bus controller or ECU (Electronic Control Unit) that sends a synchronization signal,

• A sensor controller having: i) Control means that receive the synchronization signal , ii) Means for sampling analogue signals, iii) Conversion means for converting the analogue signal into a plurality of binary signals, and particularly a sequence of binary signals, iv) Low-pass filter means for applying a low-pass filter to the plurality of binary signals, and v) Means for producing a digital value from the plurality of binary signals, wherein the control means transmit the digital value on the communication system.

In this case, the control means may be implemented (i) in hardwired form ("hardware") or as a program ("software", "firmware") . The means for sampling analogue signals (ii) , the conversion means (iii) , the low-pass filter means (iv) and the means for producing a digital value (v) may equally be hardwired - e.g. a hardwired A/D converter - or may be programmable hardware or programmable special hardware or a combination of these.

The interaction of bus controller, sensor controller ( s ) and further devices - e.g. sensor and power supply - is determined by the chosen communication protocol in conjunction with the method explained above.

In one embodiment, the low-pass filter means have a cut- off frequency that is particularly approximately 800 Hz or approximately 1600 Hz and that is correlated with the number of transmissions of the digital value by the control means after a synchronization signal. Again, it holds that the cut-off frequency of the low- pass filter means in a communication system according to the invention having periodically repeating synchronization signals with a period T_p can preferably be chosen such that said cut-off frequency is greater than 1/(2*T_P), is preferably equal to a/(2*T_p), where N > 1. At the same time, a can correspond to the number of transmissions N within the period of time T_p.

The fundamental concept in this case is that it is advantageous to increase the cut-off frequency of the low-pass filter means. This allows values to be obtained from the sensor more frequently, and the group delay time of the filter is reduced. Hence, the invention's advantages of faster reaction time and better temporal resolution of the time profile of the values are achieved . So that the advantages according to the invention can be used in the whole system, these values also need to be made known to the system and processed, i.e. the values obtained also need to be transmitted more frequently to the coordinating system component. This means that with a higher sampling rate - as a result of the higher cut-off frequency of the low-pass filter means - for an apparatus according to the invention it is necessary for the number of transmissions of the digital value to be raised.

In one preferred embodiment, the communication system is a real-time system. In a system of this kind, the use of the method according to the invention is particularly advantageous because it results in both the values that can be used becoming more precise and the reaction times becoming faster.

In a particularly preferred embodiment, the means for converting the analogue signals into a digital value is a sigma/delta analogue-to-digital converter (sigma/delta ADC) . This allows the implementation of the whole system to become more robust and less expensive. Further features of the invention will emerge from the detailed description of exemplary embodiments of the invention below in conjunction with the drawings and the claims . In the drawings:

Fig. 1: shows the schematic illustration of an apparatus according to the invention; Fig. 2: shows a comparison of different frequencies and sampling rates; shows an example of the effect of different cut-off frequencies of a low-pass filter; shows an example of the time profile in an apparatus according to the prior art with one subscriber, with a period T_p for the synchronization signal; shows an example of the time profile in an apparatus according to the invention with one subscriber, with a period T_p/2 for the synchronization signal; shows an example of the time profile in an apparatus according to the invention with two subscribers on two communication channels, with a period T_p/2 for subscriber 1 and a period T_p for subscriber 2; shows an example of the time profile in an apparatus according to the invention with two subscribers on a communication device ("dual channel device"), with a period T_p for subscriber 1 and a period T_p/2 for subscriber 2.

Fig. 1 schematically depicts the components of a - highly simplified - apparatus 100 according to the invention. In this case, the communication channel 140 is represented as a bus. This is not intended to mean that this apparatus 100 can be applied only to bus topologies; rather, the fundamental concept of the invention can also be applied to star, ring or further communication topologies. The bus 140 has a bus controller or ECU (Electronic Control Unit) 150 connected to it via the lines 151 as a central controller. This bus controller 150 undertakes central tasks in the control of the communication channel 140. Inter alia, the bus controller 150 sends a synchronization signal. In addition, the lines or line bundles 141, 142, 143 are used to connect subscribers. One of the subscribers is the sensor controller 110. This has a controller 115 that besides other tasks - receives the synchronization signal and buffer-stores the digital value, which is produced using the method described above, and transmits (sends) it on the bus 140. In addition, the controller 115 controls and monitors the execution of the method according to the invention on the sensor controller 110. To this end, the sensor controller 110 additionally has a sensor 111 that is used to sample analogue signals. In this case, the type of analogue signals is arbitrary in principle; a change in electrical values can be observed at the output of the sensor 111. A sensor according to the invention can be an acceleration sensor in a car - what is known as a crash sensor - which particularly records sharp changes in the acceleration, as an indicator of an accident, and uses this information to control the triggering of the airbags.

The analogue electrical output values from the sensor 111 are converted into a sequence of binary signals with a length of one bit in a sigma/delta modulator 112. Arranged downstream of this is a digital low-pass filter 113. This is used to determine the sampling rate, i.e. how frequently digital values are transmitted on the bus and, optionally, written to the buffer store 114 beforehand. By way of example, this buffer store 114 can also have a timestamp inserted into it, which contains information about when the digital value was sampled by the sensor 111. There may also be (not shown in Fig. 1) a second buffer store that acts as a send buffer for the bus transmission. It is possible for conversion of the digital value in the buffer store 114 to the value in the send buffer to take place.

Fig. 2 explains the effects of various sampling rates on the capture of an analogue signal. To this end, two sinusoidal curves 201 and 202 are shown. In this case, curve 201 is at the frequency 1/T_C, and curve 202 is at the frequency 2/T_c. In addition, two sampling rates are shown. The sampling instants T_c/4 are marked by circles

211, and sampling instants T_c/8 are marked by X crosses

212. It becomes clear that the sampling instants 211 allow good capture of the curve 201; the sampling instants 212 allow good capture of the curve 202.

In these cases, the sampling rate is in each case twice as high as the frequency of the sampled analogue signal ("sampling theorem", according to Shannon and Nyquist) . If one wanted to attempt to capture the curve 202 only using the sampling instants 211, then incorrect information would be supplied: the zero crossings at the instants ti and t3 are thus captured correctly, but the maximum amplitude of 202 at the instant t2 is not. In addition, it becomes clear that an arbitrary threshold 220, for a starting instant to, is reached more quickly for the curve 202 than for curve 201.

Fig. 3 schematically shows the effect of different cut- off frequencies of a low-pass filter on the reaction time of a system. To this end, an unfiltered analogue signal profile, having the reference symbol 300, is shown, along with a signal profile 301 with a low-pass filter having a low cut-off frequency and a signal profile 302 having twice the cut-off frequency. The sampling instants are marked by circles 311 (sampling rate T_c/4) and by X crosses 312 (sampling rate T_c/8), respectively.

Signal profile 300 rises sharply shortly before ti. At the instant ti, all the signal profiles 300, 301, 302 are still below the threshold 320. Before the instant t2, 300 has already reached its maximum, and 301 is just above the threshold 320, but is not captured by 311. 302 is above the threshold 320 at t2 and is captured by 312. The circles 311, on the other hand, capture the exceeding of the threshold 320 only at the instant t3. This clearly shows the effect of the faster sampling rate.

Fig. 4 shows an example of time profiles for a communication apparatus according to the prior art. An analogue signal (not shown) on the timescale 400 is sampled at the instants ti and t3, reference symbols 401 and 403. The sampling instants are each situated shortly after the pulses of the synchronization signal 411 that is sent by the bus controller 150 at the interval of time T_p. The timescale of the bus controller 410 additionally shows the timeslots 1 to 3, which each begin and end at a defined instant after the synchronization signal 411 and which are associated with the subscribers 1 to 3. After the analogue signal 400 has been measured at the instant ti and then converted into a digital value, the sensor controller 110 of subscriber 1 sends this digital value in the timeslot 1 as a data packet "Data 1" 421 via the bus. The data packet "Data 1" could also, as an alternative to the embodiment shown, be sent in the timeslot 2 or in the timeslot 3 (not shown in this case) and the number of timeslots may also be different from the three shown.

Fig. 5 shows a communication apparatus according to the invention with one subscriber having a sampling rate of twice the level. Hence, the analogue signal on the timescale 500 can be measured not only at the instants ti and t3 but also at an additional instant t2 · In this case, t2 has the period T_p/2. By applying the method according to the invention, the sensor controller 110 of subscriber 1 can send twice between two synchronization signals 511, namely data 521 in timeslot 1 and data 523 in timeslot 3. This results in the advantage that - with exactly the same hardware in the communication system - a faster sampling rate, a transmission rate matched thereto on the bus and hence a faster reaction can be achieved.

Fig. 6 shows an embodiment according to the invention with two sensor controllers; in this case, subscriber 1 has a sampling rate T_p/2 and subscriber 2 has a (different) sampling rate T_p. Timescale 620 shows the time profile of the sensor controller of subscriber 1, 630 shows the time profile of the sensor controller of subscriber 2. In this case, the sensor controller of subscriber 1 - as in Fig. 5 - sends twice, namely data packets 621 and 623, in timeslots 1 and 3, respectively.

Unaffected by this, the sensor controller of subscriber 2 can send the data packet 632 in timeslot 2. It thus becomes clear that the method according to the invention can cooperate with existing standard hardware that uses exclusively the standard protocol with no problems whatsoever .

Fig. 7 shows a further embodiment according to the invention with two subscribers on a communication device ("dual channel device") . As in Fig. 6, subscriber 1 has a sampling rate T_p/2 and subscriber 2 has a sampling rate T_p. The communication device with two "logical" sensor controllers sends on just one line bundle, which means that the time profile is shown on a single scale 720. It is therefore clearly shown that the standard protocol and the method according to the invention can also be used on the same physical line or in the same component. List of reference symbols:

100 Communication system

110 Sensor controller

111 Sensor

112 Sigma/delta modulator

113 Digital low-pass filter

114 Buffer store

115 Sensor controller control

140 Communication channel, bus

141 Bus subscriber 1

142 Bus subscriber 2

143 Bus subscriber 3

150 Bus controller, ECU

201 Signal profile 1

202 Signal profile 2

211 Circles

212 X crosses

220 Threshold

300 Signal profile 1 (analogue)

301 Signal profile 2 (filter 1)

302 Signal profile 3 (filter 2)

311 Circles

312 X crosses

320 Threshold

400 Timescale of analogue signal

401,403 Sampling instant ti, t3

410 Timescale of bus controller

411 Synchronization signal

420 Timescale of sensor controller

421 Data packet

500 Timescale of analogue signal

501,502,503 Sampling instant ti, t2, t3

510 Timescale of bus controller

511 Synchronization signal Timescale of sensor controller

, 523 Data packets

Timescale of analogue signal

, 602, 603 Sampling instant ti, t2, t3

Timescale of bus controller

Synchronization signal

Timescale of first sensor controller, 623 Data packets

Timescale of second sensor controller

Data packets

Timescale of analogue signal

,702,703 Sampling instant ti, t2, t3

Timescale of bus controller

Synchronization signal

Timescale of first sensor controller,722,723 Data packets

Claims

Method for transmitting data in a communication system (100), having the steps of:

a) sending a first synchronization signal (411, 511, 611, 711) from a bus controller (150) to at least one sensor controller (110);

b) detecting an analogue signal by means of a sensor (111) of the sensor controller (110);

c) capturing at least one signal value of the detected analogue signal at a predefined instant after the first synchronization signal (411, 511, 611, 711), as a reaction to the first synchronization signal (411, 511, 611, 711);

d) converting the at least one signal value into a plurality of binary signals;

e) filtering the plurality of binary signals by means of a low-pass filter (113) having a predefined cut-off frequency in order to produce a digital value; and

f) transmitting the digital value on a communication system, particularly on a communication bus;

characterized in that

steps b) to f) are performed N > 1 times, as a reaction to step a) .

Method according to Claim 1,

characterized in that

the plurality of binary signals and/or the digital value is stored prior to transmission thereof on the communication system in a buffer store (114) of the sensor controller (110).

Method according to Claim 1 or 2,

characterized in that

steps b) to f) are performed N > 1 times, as a reaction to step a) , before a second synchronization signal (411, 511, 611, 711) is sent.

Method according to one of the preceding claims, characterized in that

the synchronization signal (411, 511, 611, 711) is sent repeatedly and, optionally, is periodic with a period T_p.

Method according to one of the preceding claims, characterized in that

the synchronization signal (411, 511, 611, 711) is sent from the bus controller (150) to a plurality of sensor controllers (110).

Method according to one of the preceding claims, characterized in that

the communication system (100) is a synchronous communication system or a communication system that is operated in a synchronous mode.

Method according to one of the preceding claims, characterized in that

the communication system (100) is operated according to the Peripheral Sensor Interface 5, PSI5, standard .

Method according to one of the preceding claims, characterized in that

the low-pass filter (113) has a cut-off frequency that correlates with the number N of transmissions according to step f) , particularly in that the cut^¬ off frequency is greater than 1/ (2*T_P) , preferably is equal to a/ (2*T_P) , where a > 1, and particularly is approximately 800 Hz or approximately 1600 Hz.

Method according to one of the preceding claims, characterized in that

the low-pass filter (113) is a second-order or higher-order low-pass filter.

Method according to one of the preceding claims, characterized in that

the digital value has a timestamp added to it, particularly a timestamp for the capture of the signal value according to step c) and, optionally, the timestamp is transmitted with the digital value in step f) .

Method according to one of the preceding claims, characterized in that

the digital value is captured a plurality of times, particularly is captured more frequently than the captured value is transmitted according to step f) , and, optionally, is captured at equidistant instants .

Method according to one of the preceding claims, characterized in that

steps b) to e) are performed by means of a sigma/delta analogue-to-digital converter

(sigma/delta ADC, 112) .

Method according to one of the preceding claims, characterized in that

the sensor (111) is selected from a group that comprises angle sensors, pressure sensors, acoustic sensors, optical sensors, magnetic sensors, particularly Hall sensors, current sensors, voltage sensors, capacitive sensors, inductive sensors, resistive sensors, speed sensors, acceleration sensors, particularly crash sensors.

Method according to one of the preceding claims, characterized in that

the sensor controller (110) is designed as a multichannel device.

Communication system (100) for transmitting data, having

a bus controller (150) that sends a synchronization signal (411, 511, 611, 711),

a sensor controller (110) having

- control means that receive the synchronization signal (411, 511, 611, 711),

- means for sampling at least one analogue signal,

- conversion means for converting the analogue signal into a plurality of binary signals,

- low-pass filter means for applying a low-pass filter to the plurality of binary signals, and

- means for producing a digital value from the plurality of binary signals,

wherein the control means transmit the digital value on the communication system,

characterized in that

the communication system (100) implements the method according to one of the preceding claims.

Communication system (100) according to Claim 15, characterized in that the low-pass filter means have a cut-off frequency that correlates with the number N of transmissions of the digital value by the control means after a synchronization signal, particularly in that the cut-off frequency is greater than 1/ (2*T_P) , preferably is equal to a/ (2*T_P) , where a > 1, and particularly is approximately 800 Hz or approximately 1600 Hz.

17. Communication system (100) according to Claim 15 or 16,

characterized in that

the communication system (100) is a real-time communication system and, optionally, a real-time communication system operating according to the

Peripheral Sensor Interface 5, PSI5, standard.

18. Communication system (100) according to one of Claims 15 to 17,

characterized in that

the conversion means have a sigma/delta analogue-to- digital converter (sigma/delta ADC, 112).

19. Use of a communication system (100) according to one of Claims 15 to 18 for transmitting data in a real^¬ time communication system in a vehicle, preferably in a land vehicle.