Method for loss reduction in communication interface

Abstract

To provide a method, a program, and a memory medium for reducing power loss in a CAN bus driver.SOLUTION: In order to reduce loss in a communication interface having different power loss depending on bits transmitted, a data format of transmission data and / or a data format of address data indicating an address of a communication participation apparatus is selected (11, 12, 13, 22, 23, 31, 32, 3) depending on a predicted number of bits with higher power loss and / or a predicted number of bits with lower power loss in the data, such that the number of bits with lower power loss increases.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a method for reducing losses in a communication interface having different loss powers depending on the bits to be transmitted, wherein the data format of the transmission data and / or the address data is selected depending on the predicted number of higher bits of the loss power and / or the predicted number of lower bits of the loss power.

Background Art

[0002] Communication interfaces for transmitting electronic bits are widespread today. Such communication interfaces are used, for example, in vehicle manufacturing, in machine and equipment manufacturing, and in automation technology.

[0003] One such communication interface is the Controller Area Network (CAN) bus standardized by the standard ISO 11898-1. The CAN bus is a field bus system in which a number of so-called field bus modules are connected to a central control mechanism via the CAN bus. Terminal devices are also connected to the field bus modules. The terminal devices connected to the field bus system are called bus participating devices. In the CAN bus, two different bits that cause different consumption currents in the bus participating device are used. The bits are divided into so-called dominant bits and recessive bits. The dominant bit is logic 0 and applies current to the bus participating device. The recessive bit is logic 1 and only reaches the central control unit when no dominant bit is transmitted.

[0004] CAN buses are used, for example, in vehicle manufacturing to connect various radar sensors. When current is applied to radar sensors as bus-connected devices based on dominant bits, this increases power loss. This loss occurs in the driver module for the CAN bus. Higher power loss requires more expensive circuitry with relatively low-loss switching components to avoid higher costs for heat dissipation or increasing overall losses. [Overview of the Initiative] [Means for solving the problem]

[0005] A method for reducing loss in communication interfaces is proposed. A communication interface connects various communication subscriber devices, such as sensors, particularly radar sensors, and other devices. This method focuses, on the one hand, on the transmission data transmitted by the communication subscriber devices, particularly measurement data transmitted by sensors, for example. On the other hand, it focuses on the address data that indicates the address (also called ID) of the communication subscriber device and / or the message of the communication subscriber device, i.e., the address of the transmitted data packet.

[0006] Communication interfaces have different power losses depending on the bits being transmitted. For example, a communication interface may have higher or lower power losses when transmitting bits with logic 1 compared to bits with logic 0. Depending on the data format of each data, it is typical that there will be more bits with higher power losses or more bits with lower power losses. Depending on the application, the predicted number of bits with higher power losses or lower power losses within each data format can be estimated. This estimation can be done during the design, fabrication, or installation of the communication interface, or later, for example, by updating the communication interface.

[0007] The selected data format is intended to increase the number of bits with lower power loss, depending on the predicted number of bits with higher and / or lower power loss in the data when applying the data format for the transmission data and / or the data format for the address data. It is preferable to maximize the number of bits with lower power loss. In other words, appropriate selection of the data format increases the number of bits with lower power loss, thereby reducing power loss during transmission.

[0008] Reducing power loss during transmission offers the following advantages: Firstly, the total power loss of the communication subscriber equipment decreases, resulting in higher quality. Secondly, the total power loss of the communication subscriber equipment can be kept constant. Thirdly, the supply voltage can be provided more easily and favorably. This can be achieved, for example, by using more suitable components with higher losses, or by changing the circuit topology, such as using a linear regulator instead of a switching regulator with higher efficiency. Finally, more suitable mechanical components, which may have worse thermal characteristics, can be used. This can reduce the cost of the communication interface.

[0009] Generally, this method can be applied to any communication interface where power loss depends on the bits being transmitted. Such a communication interface could be a fieldbus, for example. In particular, this communication interface is the CAN bus, which has dominant bits (logical 0s) and recessive bits (logical 1s). Dominant bits are bits with higher power loss, and recessive bits are bits with lower power loss. The CAN bus connects various bus subscribers to a central control mechanism. These bus subscribers are the communication subscribers described above. Depending on the data format of each data, it is typical that the number of dominant or recessive bits increases. For example, integers with a large word width (and therefore many bits) and low values ​​have many leading 0s, and therefore many dominant bits.

[0010] A method for reducing losses in the CAN bus is to select the data format for the transmission data and / or address data, depending on the predicted number of dominant bits and / or recessive bits in the data, so that the selected data format increases the number of recessive bits in the data. It is preferable to maximize the number of recessive bits. Recessive bits consume less current than dominant bits; for example, according to datasheet TCAN1042-Q1, the current consumption is reduced by 75mA. Therefore, increasing the number of recessive bits and simultaneously decreasing the number of dominant bits during transmission reduces the total current. Consequently, the power loss generated during transmission is reduced; in this case, at a voltage of 5V, the power loss is reduced by 375mW.

[0011] If the transmitted data is, for example, sensor measurement data, it is preferable that the data format of the measurement data be selected depending on the type of measurement and / or the type of evaluation of the measurement data. For example, the data format may be selected depending on whether the measurement data is used to determine distance or relative velocity. Alternatively, or in addition to the above, the data format of the measurement data may be selected depending on the location of the bus-joining equipment, i.e., the sensor.

[0012] Distance is typically output as an unsigned integer on a fixed scale. Conventional data formats assign lower values ​​to shorter distances and higher values ​​to longer distances. In distance measurement, it is typical that more targets within a relatively close vicinity are captured than distant targets. Therefore, in conventional data formats, many close targets are assigned low values, and these low values ​​have many leading zeros in binary representation. This results in many bits with higher power loss during transmission, where logical zeros correspond to bits with higher power loss, as is the case with dominant bits in CAN buses. Therefore, according to the present invention, a data format that inverts the bits before transmission may be selected. Thus, leading zeros become leading ones, and one in this case corresponds to a bit with lower power loss, as is the case with recessive bits in CAN buses.

[0013] Signed integers are used in the measurement of relative velocity. For front sensors pointed in the direction of movement, most targets, especially all stationary targets whose position does not change while the sensor is moving toward them, typically have a negative relative velocity. For rear sensors pointed in the opposite direction of movement, most targets typically have a positive relative velocity. Conventionally, one's complement representation is used for relative velocity measurement data. This data format is already optimal for front sensor measurements, which primarily measure negatively signified relative velocities, where logical 0 corresponds to a higher bit of power loss. For rear sensor measurements, which primarily measure positively signified relative velocities, the present invention may select a data format inverting the bits before transmission. Thus, a different data format is selected depending on the sensor's position, specifically whether it is front or rear. When the direction of movement changes, for example during reverse driving, it may be intended to prevent the rear sensor measurement data from being inverted.

[0014] It is preferable that the selected data format is communicated to the electronic control unit that controls the communication interface. In the case of a CAN bus, this is the central control unit. For this purpose, an additional bit may be intended during transmission, for example, to indicate inversion. This allows the electronic control unit to correctly interpret the measurement data. This is particularly noteworthy when the measurement data is inverted or not, depending on the position and, possibly, the direction of movement, as mentioned above. Generally, multiple bits may also be used to indicate the data format. In particular, when only one bit is used to indicate the data format, the number of bits converted from higher power loss bits to lower power loss bits by changing the data format is usually significantly greater than the number of bits used to indicate the data format, because the number of detections to be transmitted is large.

[0015] If the data format intends to have a default value for unused values ​​in the data packet, it is preferable that the default value is selected depending on the predicted number of bits with higher and / or lower power loss in the data, such that, in the case of a default value, the number of bits with lower power loss increases in the selected data format. For example, in distance measurement using a sensor, all detection data is always transmitted regardless of how many detections the sensor actually captures. A default value can then be determined for detections that were not captured. Simply as an example, such a default value may consist only of bits with lower power loss. In the case of a CAN bus, the default value may consist only of 1, for example, and therefore only of recessive bits.

[0016] When a data compression method using a dictionary, such as Huffman coding, is used during data transmission, it is preferable that this dictionary is selected depending on the predicted number of bits with higher power loss and / or lower power loss in the data, such that the number of bits with lower power loss increases. In this case, the frequency distribution of words is taken into consideration in particular, and when implemented, the frequency of bits with higher power loss decreases. In this regard, it is intended that the amount of data does not decrease when the number of bits with higher power loss decreases, or even increases, only if the system as a whole allows.

[0017] With regard to address data, it is preferable that the addresses / IDs of bus subscriber devices be selected based on an assignment table. The assignment table assigns a unique address / ID to each subscriber device. Alternatively, or in addition to this, the addresses / IDs of subscriber device messages, i.e., transmitted data packets, may be selected based on the assignment table. In this case, the assignment table assigns a unique address / ID to each subscriber device's various messages or data packets. In this regard, it is preferable that the assignment table uses only values ​​for addresses / IDs that contain as few higher-power-losing bits as possible in binary representation, and particularly preferable that it uses only values ​​that contain at most one higher-power-losing bit in binary representation (so-called 1-of-N encoding or one-hot encoding).

[0018] The computer program is configured to perform each step of this method, particularly when implemented on electronically controlled devices of communication interfaces. The computer program enables the implementation of this method on conventional electronically controlled devices without requiring any structural changes. For this purpose, the computer program is stored on a machine-readable memory medium.

[0019] By installing this computer program in a conventional electronic control device of a communication interface, the data format of the transmission data and / or the data format of the address data are selected depending on the predicted number of higher-loss bits and / or the predicted number of lower-loss bits in the data, such that the number of lower-loss bits overwhelms, and thus reduces the loss power of the communication interface. An electronic control device is obtained in which this is set.

[0020] Exemplary embodiments of the present invention are shown in the drawings and are explained in more detail in the following description.

Brief Description of the Drawings

[0021] [Figure 1] It is a schematic diagram of a vehicle equipped with a radar sensor and a CAN bus connected via the radar sensor, and a detected target. [Figure 2] It is a flowchart of an embodiment of the method according to the present invention. [Figure 3] It is a diagram showing an allocation table for addresses based on the prior art. [Figure 4] It is a diagram showing an allocation table for addresses based on an embodiment of the method according to the present invention.

Modes for Carrying Out the Invention

[0022] Figure 1 shows a schematic view of a vehicle FZ moving in the driving direction FR. This vehicle has a CAN bus CAN and, by way of example, eight radar sensors S1 to S8. The eight radar sensors S1 to S8 are connected to a central control unit ZSE in the middle of the CAN bus CAN via the CAN bus CAN. In addition to this, by way of example, targets Z1 to Z3 detected by the radar sensors S1 to S8 are shown. In this example, the front sensors S1 to S3 facing the driving direction FR capture the targets Z1 & Z2 in front of the vehicle FZ, and the rear sensors S6 to S8 facing the opposite side capture the target Z3 behind the vehicle FZ. The radar sensors S1 to S8 can perform, on the one hand, the measurement of the distance to the targets Z1 to Z3 and, on the other hand, the measurement of the relative speed of the targets Z1 to Z3 with respect to the vehicle FZ.

[0023] Figure 2 shows a flowchart of an embodiment of the method according to the invention. On the one hand, the data format for the measurement data of the radar sensors S1 to S8 is changed (left path), which will be first explained below. On the other hand, the data format of the address data is changed (right path), which will be explained subsequently with reference to Figures 3 and 4 as well.

[0024] First of all, an inquiry 1 is made as to whether the measurement data is recorded during a distance measurement (displayed as "d") or during a measurement of the relative speed (displayed as "V rel "). In the case of a distance measurement, measurement 10 is carried out by all the radar sensors S1 to S8, during which the distance to each of the targets Z1 to Z3 is measured. The measurement data obtained in measurement 10 exists as unsigned integers on a fixed scale respectively assigned to the distances. Conventionally, lower numerical values are assigned to smaller distances and higher numerical values are assigned to larger distances. That is, for example, a distance of 31.25 m for target Z1 corresponds to the number 8192 (decimal), and a distance of 250 m for target Z2 corresponds to the number 65535 (decimal). Based on the attenuation of the radar wave, the detection frequency of smaller distances is higher than that for larger distances. However, smaller distances have lower numerical values, and lower numerical values have many leading 0s in binary representation, thus generating many dominant bits during transmission.

[0025] Therefore, the data format of the measurement data is changed according to the present invention. For this purpose, the measurement data is inverted 11. This results in numerical values ​​for small distances having many leading 1s instead of leading 0s, thus generating more recessive bits during transmission. In addition, a default value is determined 12 for unused values. Regardless of how many detections the radar sensor actually measures, 128 detections are always transmitted. The default value for unused values ​​may include, for example, only 1, which generates only recessive bits during transmission. The default value can always be predetermined 12. In this exemplary embodiment, in addition to this, compression using a dictionary 13, such as Huffman coding, is performed. The dictionary is selected for compression 13 so as to reduce the frequency of dominant bits.

[0026] Finally, the measurement data in the modified data format is transmitted to the central control unit ZSE via the CAN bus CAN 14. Simultaneously, information about the modified data format may also be transmitted to the central control unit ZSE via the CAN bus CAN 15. In the simplest case, the inverted measurement data 11 is presented as one bit. The central control unit ZSE can now process the measurement data in the modified data format as well.

[0027] Relative velocity V relIn the measurement, the first step is to inquire about the location of the sensors to be measured, that is, whether the measurement is performed by the front sensors S1-S3, the side sensors S4 & S5, or the rear sensors S6-S8. If the measurement 20 is performed by the front sensors S1-S3, the relative speed of most targets Z1 and Z2 in front of the vehicle FZ is negative, based on the speed of the vehicle FZ in the direction of travel FR, especially in the case of stationary targets, which are the most frequently occurring. The measurement data obtained in measurement 20 conventionally exists as a signed integer, represented using one's complement. This measurement data is already inverted in the case of one's complement due to the negative sign. Therefore, this measurement data has many leading 1s, which results in many recessive bits during transmission. In this case, only standard values ​​are determined for the values ​​that are not used. See the explanation above for details.

[0028] When measurements 30 are performed by the rear sensors S6~S8, the relative speed of most targets Z3 behind the vehicle FZ is positive, based on the speed of the vehicle FZ in the direction of travel FR, especially in the case of stationary targets, which are the most frequent occurrences. The measurement data obtained in measurement 30 also conventionally exists as a signed integer, represented in one's complement. However, because this measurement data has a positive sign, it is not inverted in one's complement. In this case, it is intended that the measurement data be inverted 31. This inversion 31 results in the measurement data having many leading ones, which generates many recessive bits during transmission. In addition, following the front sensor, a standard value is determined 32 for values ​​that are not used. See the explanation above for details.

[0029] In the case of side sensors S4 and S5, targets Z1 to Z3 may be located in front of or behind the vehicle FZ, or they may move from one to the other as they pass by, so there is no preferred direction for relative speed. With respect to the method according to the present invention, the side sensors can be treated in principle like front sensors or like rear sensors. In the embodiment illustrated here, side sensors S4 and S5 are treated like front sensors and measurement 20 is performed. In this case, no inversion of the measurement data is performed.

[0030] Generally, the order of measurement 20 or 30 and query 2 is not predetermined. That is, query 2 may be performed after measurement 20 or 30 to determine which sensors S1-S3, S4 & S5, or S6-S8 the measurement is for. In addition, standard values ​​can always be predetermined 22 or 32.

[0031] Finally, the measurement data in the modified data format is transmitted to the central control unit ZSE via the CAN bus CAN 24. Simultaneously, information about the modified data format is transmitted to the central control unit ZSE via the CAN bus CAN 25. In the simplest case, the inversion 31 of the measurement data is presented as a single bit. This allows the central control unit ZSE to distinguish whether the measurement data is from the front sensors S1~S3 (and side sensors S4&S5) and is therefore inverted based on one's complement and has a negative sign, or whether it is from the rear sensors S6~S8 and is therefore inverted 31 according to the method according to the present invention and has a positive sign, thus the central control unit ZSE can now process the measurement data in the modified data format as well.

[0032] In other exemplary embodiments not presented, firstly, the distance d and relative velocity V of all radar sensors S1-S8 are measured. rel The same measurement is then performed. In this case, it is then determined which part of the measurement data will be used to determine the distance d and which part will be used to determine the relative velocity V. relQuery 1 is made to determine what will be used to determine the type of measurement data to evaluate. Therefore, measurements 10, 20, and 30 are eliminated, and instead, the corresponding portions of the measurement data are used. For example, in response to query 2 for the sensors to measure, portions of the measurement data are selected for each sensor.

[0033] Figures 3 and 4 show assignment tables for addresses (also called IDs) of bus subscriber devices, i.e., sensors S1-S8 (also referred to here as S). Each table lists the sensor S, the address ID of sensor S as a decimal number (indicated as "dez") and an 11-digit binary number (indicated as "bin"), and the number of zeros N (0) in the binary number. Figure 4 shows an assignment table used for assigning address IDs to bus subscriber devices 3 (see Figure 2) according to the method of the present invention. Figure 3 shows a conventional assignment table based on the prior art for comparison. In the conventional assignment table, the address IDs are assigned to the sensors S in ascending order starting from the digit 0 (dez) in the simplest case. This results in a total of 76 zeros, and therefore dominant bits, for the eight sensors S1-S8. In contrast, if the address IDs are assigned to the sensors S in descending order starting from the digit 2047 (dez), this reduces the total number of zeros for the eight sensors S1-S8 to 12.

[0034] Figure 4 shows a preferred assignment table using 1-of-N encoding (also known as one-hot encoding). Here, the address ID is assigned to the sensor S such that each of its binary representation values ​​contains at most one zero, with each zero being moved one digit forward. For the eight sensors S1 to S8, the total number of zeros is 7.

[0035] An allocation table for the addresses of messages from bus subscriber devices, i.e., the data packets being transmitted, may also be provided.

[0036] The following illustrates loss reduction based on illustrative CAN bus figures. It should be added that in practice, variations and uncertainties will occur depending on the bus used, temperature, and / or other chip conditions. See also datasheet TCAN1042-Q1 for further details. The CAN bus transmission rate (also called bus speed) averages 1 Mbps, i.e., 10 6 The rate is bits per second. Therefore, the time it takes to apply one bit to the bus is 1 μs. The power loss per faulty bit is 0.375 μW, assuming this bit is transmitted at a rate of 1 bit per second.

number

[0037] - If we improve by 1 bit per measurement, this becomes 20 bits / second, and therefore 7.5 μW.

number

number

number

number

number

[0038] CAN CAN bus FR (Front-engine, Rear-wheel drive) FZ Vehicle S1~S8 Communication equipment Z1~Z3 Goal ZSE Central Control Unit d Distance measurement V rel Relative velocity 1. Inquiry about the type of measurement 2. Inquiry about the location of the sensor to be measured. 3. Choices 10 measurements 11 Reversal 12 decision 13. Data compression using dictionaries 14 Transmission 15. Transmission of Information 20 measurements 22 Decision 23. Data compression using dictionaries 24 Transmission 25. Information Transmission 30 measurements 31 Reversal 32 Decision

Claims

1. 1. A method for reducing losses in a communication interface having different power losses depending on the bits to be transmitted, characterized in that the data format of the transmitted data and / or the data format of the address data indicating the addresses (IDs) of the communication subscribers (S1-S8) and / or messages of the communication subscribers (S1-S8) are selected (11, 12, 13, 22, 23, 31, 32, 3) depending on the expected number of higher power loss bits and / or the expected number of lower power loss bits in said data, such that the number of lower power loss bits is increased.

2. 2. The method of claim 1, wherein the communication interface is a CAN bus (CAN), and the dominant bits are bits with higher power dissipation and the recessive bits are bits with lower power dissipation.

3. 2. The method according to claim 1, wherein the transmitted data are measurement data, and the data format of the measurement data is selected (11, 31) depending on the type of measurement and / or the type of evaluation of the measurement data and / or the positions of the bus participants.

4. 2. A method according to claim 1, characterized in that the selected data format is communicated (15, 25) to an electronic control device controlling the communication interface.

5. 2. The method of claim 1, wherein standard values for unused values in the data format are selected (12, 22, 32) depending on the expected number of higher power dissipation bits and / or the expected number of lower power dissipation bits in the data, such that the number of lower power dissipation bits increases.

6. 2. The method according to claim 1, characterized in that in a data compression method (13, 23) using a dictionary, the dictionary is selected depending on the predicted number of higher power dissipation bits and / or the predicted number of lower power dissipation bits in the data, so as to increase the number of lower power dissipation bits.

7. 2. A method according to claim 1, characterized in that the address (ID) of the communication participant is selected (3) on the basis of an allocation table.

8. 2. The method of claim 1, wherein the address of the message of the communication participant is selected based on an allocation table.

9. 8. The method according to claim 7, characterized in that for the addresses (ID) an allocation table is used which uses only values which in their binary representation contain at most one higher bit of dissipated power.

10. A computer program configured to carry out the steps of the method according to any one of claims 1 to 9.

11. A machine-readable memory medium having stored thereon the computer program of claim 10.

12. 10. An electronic control device that is configured for selecting a data format for transmission data and / or address data in the case of a CAN bus (CAN) using a method according to any one of claims 1 to 9.

Images