Vehicle and procedures for in-vehicle communication

Abstract

Vehicle (100), comprising: a sending control unit (110-T) configured to send a message; a receiving control unit (110-R) configured to receive the message; one or more switches (120) configured to forward the message from the sending control unit (110-T) to the receiving control unit (110-R), wherein the sending control unit (110-T) and each of the one or more switches (120) include an output port for transmitting the message; and a processor configured to perform limited timing analysis to determine a total wait time w k during the transmission of the message from the sending control unit (110-T) to the receiving control unit (110-R) as a sum of all output port wait times w i,jto determine at all outgoing ports to which the message encounters, taking measures to avoid or mitigate the total waiting time w k during transmission a time limit for the message is exceeded, and the limited timing analysis includes performing an iterative process and determining a lower bound (LB), an upper bound (UB), and a median value.

Description

[0001] The present disclosure relates to a limited timing analysis of vehicle-internal communication.

[0002] A vehicle (e.g., automobile, truck, construction equipment, agricultural machinery, automated factory equipment) typically includes multiple control units referred to as electronic control units (ECUs). The ECUs can communicate with each other in a multi-hop network configuration, where a message from a sending ECU to a receiving ECU may involve the use of one or more intermediate Ethernet switches to relay the message. As a result, the queuing delay a message experiences at each output port associated with the sending ECU and each intermediate switch can complicate timing analysis (i.e., determining the total delay for communicating the message). Accordingly, it is desirable to provide limited timing analysis of the vehicle's internal communication.

[0003] US 2007 / 0286097A1 relates to a system and method for self-organizing, reliable data transmission over multiple paths in a network, which uses queues to transfer messages between end-user modules (EUMs) on nodes in the network.

[0004] The objective can be considered to be to provide a vehicle and an alternative method for improved timing analysis for in-vehicle communication. This objective is achieved by the subject matter of claim 1 and claim 6.

[0005] The vehicle according to the invention comprises a sending control unit for transmitting a message and a receiving control unit configured to receive the message. The vehicle also includes one or more switches configured to forward the message from the sending control unit to the receiving control unit. The sending control unit and each of the one or more switches include an output port for transmitting the message. A processor performs limited timing analysis to determine the total wait time during the transmission of the message from the sending control unit to the receiving control unit as the sum of all output port wait times at all output ports to which the message arrives.Measures are taken to avoid or mitigate the total waiting time during transmission exceeding a message deadline, and the limited timing analysis includes performing an iterative process and determining a lower bound (LB), upper bound (UB), and median value.

[0006] According to one embodiment, the processor is part of the sending control unit, the receiving control unit, or a third control unit.

[0007] According to one embodiment, the processor determines an initial value for the lower limit (LB) and for the upper limit (UB).

[0008] According to one embodiment, the processor is configured to calculate the median value at each iteration of the iterative process as follows: median=(LB+UB)2.

[0009] According to one embodiment, the processor calculates an initial value for the output port wait time w at each iteration of the iterative process. i,j at a given j-th of the output ports as follows: wi,j=α+∑⌈medianTi⌉Ci,where i is an index for the message, T i a period of the message is and C i a transmission time for each transmission of the message.

[0010] According to one embodiment, the output port wait time w i,j The message is determined as the median value based on the fact that the median value calculated for the iteration is equal to the median value for the output port wait time w. i,j The calculated initial value for the iteration is...

[0011] According to one embodiment, the initial value for the lower bound (LB) or for the upper bound (UB) is adjusted before the next iteration of the iterative process based on the fact that the median value calculated for the iteration is not equal to the median value for the output port wait time w. i,j The calculated initial value for the iteration is...

[0012] According to one embodiment, the measure to prevent the total waiting time during transmission from exceeding the time limit for the message includes discarding another message that precedes the message in a queue at one of the one or more switches for forwarding the message.

[0013] According to one embodiment, the measure to mitigate the fact that the total waiting time during a transmission exceeds the message deadline includes reducing the service level of a service of the vehicle provided by communicating the message.

[0014] According to one embodiment, the receiving control unit controls the operation of the vehicle based on the message.

[0015] The inventive method for in-vehicle communication in a vehicle comprises sending a message from a sending control unit of the vehicle to a receiving control unit and forwarding the message between the sending and receiving control units using one or more switches. The sending control unit and each of the one or more switches include an output port for transmitting the message. The method also includes performing a limited timing analysis to determine the total waiting time during the transmission of the message from the sending control unit to the receiving control unit as the sum of all output port waiting times at all output ports to which the message arrives.Performing limited timing analysis involves an iterative process to determine a lower bound (LB), an upper bound (UB), and a median value. Measures are taken to prevent or mitigate the total waiting time during transmission from exceeding a message deadline.

[0016] According to one embodiment, the limited timing analysis is performed by the sending control unit, the receiving control unit, or a third control unit.

[0017] According to one embodiment, the method also includes determining an initial value for the lower limit (LB) and for the upper limit (UB).

[0018] According to one embodiment, the method also includes calculating the median value at each iteration of the iterative process as: median=(LB+UB)2.

[0019] According to one embodiment, the method also includes calculating an initial value for the output port wait time w. i,j at a given j-th of the output ports at each iteration of the iterative process, the following applies: wi,j=α+∑⌈medianTi⌉Ci,where i is an index for the message, T i a period of the message is and C i a transmission time for each transmission of the message.

[0020] According to one embodiment, the method also includes determining the output port wait time w. i,j the message is entered as the median value based on the fact that the median value calculated for the iteration is equal to the median value for the output port wait time w. i,j The calculated initial value for the iteration is...

[0021] According to one embodiment, the method also includes adjusting the initial value for the lower bound (LB) or for the upper bound (UB) before the next iteration of the iterative process based on the fact that the median value calculated for the iteration is not equal to the median value for the output port wait time w. i,j The calculated initial value for the iteration is...

[0022] According to one embodiment, taking the measure to prevent the total waiting time during transmission from exceeding the message deadline includes discarding another message that precedes the message in a queue at one of the one or more switches configured to forward the message.

[0023] According to one embodiment, taking the measure to mitigate the fact that the total waiting time during a transmission exceeds the message deadline includes reducing the service level of a service of the vehicle provided by communicating the message.

[0024] According to one embodiment, the method also includes the receiving control unit controlling the operation of the vehicle based on the message.

[0025] The above features and advantages, as well as other features and advantages of the disclosure, will be readily apparent from the following detailed description in conjunction with the accompanying drawings.

[0026] Other features, advantages, and details appear only as examples in the following detailed description, which refers to the drawings in which: Fig. 1 a block diagram of a vehicle that includes systems for performing a limited timing analysis of the vehicle's internal communication according to one or more embodiments; Fig. 2 illustrates an exemplary vehicle-internal communication for which a limited timing analysis is performed according to one or more embodiments; Fig. 3 shows a process flow of a method for performing a limited timing analysis of the vehicle's internal communication according to one or more embodiments; and Fig. 4 shows an exemplary traffic flow to discuss aspects of the limited timing analysis of vehicle-internal communication according to one or more embodiments.

[0027] The following description is merely exemplary and is not intended to limit the present disclosure, its application, or uses. It should be understood that throughout the drawings, corresponding reference numerals indicate identical or corresponding parts and features.

[0028] As previously noted, a vehicle can include multiple ECUs that communicate with each other. These ECUs can be used, for example, for autonomous vehicle operation or for advanced vehicle systems (such as collision avoidance, adaptive cruise control, and automatic braking). To control one or more vehicle operations, two or more ECUs can communicate with each other. As mentioned earlier, communication between ECUs in a vehicle can occur via a multi-hop network, which involves the use of Ethernet switches to relay a message. Therefore, the overall time required to deliver a message must account for the queuing delay associated with the sending ECU and each Ethernet switch between the sending and receiving ECUs.

[0029] Many messages exchanged by ECUs can have deadlines, so determining the end-to-end latency (i.e., the total time the message takes to travel between the sending and receiving ECUs) is an important aspect in determining whether the deadline is met. If a deadline has been missed or will not be met, the sending ECU or another control unit can take one of several actions. The action may depend on the type of vehicle operation (for example, infotainment system, safety system) to which the message belongs. One example action is to notify an operator (for example, the driver) that the service associated with the message has degraded or is unavailable. This action may be appropriate for messages related to infotainment or other non-critical systems.Another exemplary measure is to discard other messages to reduce the latency associated with delivering the specific message. This measure may be suitable for a relatively high-priority message. Prior art approaches for determining end-to-end latency (i.e., performing timing analysis) involve the use of a recursive equation. Thus, limiting the execution time for performing timing analysis is difficult. The embodiments of the systems and methods detailed here relate to a limited timing analysis of in-vehicle communication. A pseudobinary search-based timing analysis technique is used, as further explained below.

[0030] According to one embodiment, Fig. 1 A block diagram of a vehicle 100, which includes systems for performing a limited timing analysis of the vehicle's internal communication. The in Fig. The vehicle 100 shown is a passenger car 101. The exemplary vehicle 100 includes a set of ECUs 110-1 to 110-n (generally referred to as 110) and Ethernet switches 120-1 to 120-x (generally referred to as 120). The vehicle 100 may also include one or more sensors 130 (e.g., a radar system, a lidar system, and a camera). One or more ECUs 110 can process data from the sensors 130. In addition, the ECUs 110 can perform autonomous driving or control various aspects of the operation of the vehicle 100 (e.g., braking, steering). While in Fig. While exemplary locations for the ECUs 101, Ethernet switches 120, and sensors 130 are specified, the location and relative arrangement of each component may differ according to alternative embodiments. The ECUs 110 may include processing logic and other components. The processing logic of the ECUs 110 may include an application-specific integrated circuit (ASIC), an electronic circuit, a (shared, dedicated, or group) processor and memory that executes one or more software or firmware programs, a combinational logic circuit, and / or other suitable components that provide the described functionality.

[0031] Fig. Figure 2 illustrates exemplary in-vehicle communication for which a limited timing analysis is performed according to one or more embodiments. The transmitting ECU 110-T and the receiving ECU 110-R are shown with three intervening Ethernet switches 120-1, 120-2, and 120-3. The ECUs 110 and Ethernet switches 120 can share a bus 210 for communication instead of being connected to each other, as shown for illustrative purposes. The message queue in Ethernet switch 120-3 is shown. The messages M-1 to Mx are arranged by priority such that M-1 is the highest-priority message at the output port of Ethernet switch 120-3, and Mx is the lowest-priority message in the message queue. The path from ECU 110-T→120-1→120-2→120-3→110-R is an example path for an example message (e.g. M-2), however, other messages may take different paths.For example, messages shown in the queue at Ethernet switch 120-3 can take different paths, so that, for example, M-1 can be sent to another Ethernet switch 120 or another ECU 110 instead of the ECU 110-R.

[0032] The total waiting time for a message (e.g., M-2) between the ECU 110-T and the ECU 110-R can be expressed as w k be designated. The total waiting time w k For a given i-th, a sum of the outbound port wait times w is given. i,j at all j output ports that the message encounters (i.e., the waiting time at ECU 110-T and at each Ethernet switch 120 between ECU 110-T and ECU 110-R). In the Fig. In the exemplary case shown in 2, the total waiting time w must be k the queue delay (i.e., the exit port wait time w i,j) at the output port of the ECU 110-T up to the queue delay at the output port of the Ethernet switch 120-3. That is, each queue delay can be considered as w i,j This can be expressed as, where i is the index for a given message and j is the index for a given output port. As noted previously, a recursive calculation was used to determine the total wait time w. k to determine. As with reference to Fig. 3 is described in detail, the process for determining the total waiting time w k limited according to one or more embodiments. In particular, an upper limit (UL) and a lower limit (LB) are used to ensure convergence for the total waiting time w. k to reach.

[0033] Fig. Figure 3 shows a process flow 300 of a method for performing a limited timing analysis of the vehicle's internal communication according to one or more embodiments. The process flow 300 can be performed at an ECU 110 that sends a message, or at a central ECU 110 that may not be involved in the communication itself but determines the latency for vehicle-internal communication. According to an alternative embodiment, the processes can also be performed by more than one ECU 110. As detailed, the processes are performed at each block, except for part of the process at block 340 and the process at block 345, for each port (i.e., each index j) traversed by the message (i.e., index i) for which the total waiting time w is determined. kis calculated. For block 310, setting initial values ​​for LB and UB includes setting the initial value of LB as the known time required to clear bus 210 and reset between messages. The initial value of the upper bound UB for the specific message (index i) and output port 120 (index j) can (for block 310) be obtained from the following equations using an existing upper bound analysis technique: UBi,j=Ci+1+b1…i1−U1…i b1…i=∑k=1iCk(1−Uk1−U1…k−1) Ui=cibus bandwidth U1…K=∑k=1iUk

[0034] In the equations, C i The transmission time for the i-th message at the j-th output port is 120. The priority of messages with index k < i is greater than the priority of message i. The bus bandwidth in GL. 3 is the bandwidth of the bus 210.

[0035] For block 320, calculating a median value and the output port wait time w is completed. i,j the implementation of the following formulas: median=(LB+UB)2 wi,j=α+∑⌈medianTi⌉Ci

[0036] In GL. 6, i is an index for the message, T i is the message period (i.e., how often the message needs to be repeated). At block 330, it is checked whether the calculated output port wait time w is correct. i,j (in GL. 6) is equal to the median (in GL. 5). If this is the case, the processes conclude that providing the median value is the output port wait time w. i,j at block 340.

[0037] If the check at block 330 shows that the calculated output port wait time w i,j If the value is not equal to the calculated median, a further check is performed at block 350 to determine if the output port wait time w i,jis smaller than the median. If the calculated output port wait time w i,j If the value is smaller than the median (for block 350), the upper limit (UB) for block 360 is set to the value of the exit port wait time w. i,j The process is set up and another iteration begins at block 320, starting with the calculation of the median value and the output port wait time w. i,j performed. If the calculated output port wait time w i,j If the median is not smaller than the median (for block 350), the last termination time θ of messages released before the median is calculated for block 370. The last termination time θ is determined with reference to Fig. 4. This is discussed further. At block 380, it is checked whether the last termination time θ is greater than the median value (calculated at block 320). If the last termination time θ is greater than the median, the upper bound (UB) at block 390 is set to the median, and at block 320, another iteration begins, starting with the calculation of the median value and the exit port wait time w. i,j , performed. If the last termination time θ is not greater than the median, the lower bound (LB) for block 385 is set to the exit port wait time w. i,j set, and at block 320 another iteration is started, beginning with the calculation of the median value and the output port wait time w. i,j carried out.

[0038] As soon as (at block 340) the exit port wait time w i,jFor all output ports (i.e., sending ECU 110 and intermediate Ethernet switches 120) between the communicating ECUs 110, a sum of the output port wait times w is obtained. i,j all output ports (i.e., all j) are calculated (at block 340) to determine the total wait time w k to obtain the message (i.e., the relevant i). In block 345, avoiding or mitigating a delay refers to measures that can be taken if the total waiting time w k does not coincide with the deadline for the given message. As previously noted, a message can have a deadline assigned to it, so that based on the total waiting time w k It can be determined whether the message will reach the receiving ECU 110 within the deadline. The deadline for a given message can, for example, be based on its priority. As also noted previously, a statement that the total waiting time w kFailure to meet the deadline will result in a notification or other action. Avoiding a delay at block 345 involves modifying the message queue (for example, discarding one or more other messages) to ensure the deadline is met. Mitigating the delay at block 345 involves reducing the level of any service provided by the message. If the message reaches the receiving ECU 110 within the deadline, as described in Fig. Based on the 3 processes shown, the message can be used to control an aspect of vehicle operation (for example, collision avoidance, automatic braking, infotainment) or the autonomous operation of the vehicle.

[0039] Fig. Figure 4 shows an exemplary traffic flow at an exemplary exit port to discuss aspects of the limited timing analysis of in-vehicle communication according to one or more embodiments. More specifically, the determination of the last termination time θ is explained using exemplary messages M1, M2, M3, and M4, which are shown in Fig. 4 shown. The table in Fig. Column 4 specifies the period Ti and transmission time Ci of each message. The example median is 5.1. The time values ​​(Ti, Ci, median) can be expressed in any unit of time, such as milliseconds (ms). The last column of the table shows the transmission time of each message if no other messages existed (i.e., if there were no queue to consider). Only transmission times preceding the median value of 5.1 are shown. Thus, for example, 7 is not shown in addition to 0 for message M4, which has a period of 7. Instead, only 0 is shown, as it precedes the median value of 5.1.

[0040] From this information, the following can be found below. Fig. The timeline shown in Figure 4 can be derived from this. Essentially, the last (i.e., bold) of the transmission times listed in the last column is the time of interest for each message, but message M2 cannot be transmitted at time 4 due to interference with the transmission of message M1, and message M3 cannot be transmitted at time 5 due to interference with the transmission of message M2. Instead, message M2 is transmitted after the transmission time of message M1, and message M3 is transmitted after the transmission time of message M2. As shown in Figure 4, the transmission times shown in Figure 4 are the last (i.e., the bold) time of each transmission time shown in the last column are the time of interest for each message. However, message M2 cannot be transmitted at time 4 due to interference with the transmission of message M1, and message M3 cannot be transmitted at time 5 due to interference with the transmission of message M2. Instead, message M2 is transmitted after the transmission time of message M1, and message M3 is transmitted after the transmission time of message M2. Fig. As shown in figure 4, all messages are transmitted up to 6.5. For this example, this is the value of the last termination time θ.

Claims

[1] Vehicle (100), comprising: a sending control unit (110-T) configured to send a message; a receiving control unit (110-R) configured to receive the message; one or more switches (120) configured to forward the message from the sending control unit (110-T) to the receiving control unit (110-R), wherein the sending control unit (110-T) and each of the one or more switches (120) include an output port for transmitting the message; and a processor configured to perform limited timing analysis to determine a total wait time w k during the transmission of the message from the sending control unit (110-T) to the receiving control unit (110-R) as a sum of all output port wait times w i,jto determine at all outbound ports to which the message encounters, taking measures to avoid or mitigate the total waiting time w k during transmission a time limit for the message is exceeded, and the limited timing analysis includes performing an iterative process and determining a lower bound (LB), an upper bound (UB), and a median value. [2] Vehicle (100) according to claim 1, wherein the processor is part of the sending control unit (110-T), the receiving control unit (110-R) or a third control unit and the receiving control unit (110-R) controls an operation of the vehicle (100) based on the message. [3] Vehicle (100) according to claim 1, wherein the processor is configured to determine an initial value for the lower bound (LB) and for the upper bound (UB), wherein the processor is configured to calculate the median value at each iteration of the iterative process as follows: median=(LB+UB)2, where the processor is configured to provide an initial value for the output port wait time w at each iteration of the iterative process i,j to be calculated at a given j-th of the output ports as follows: wi,j=α+∑⌈medianTi⌉Ci,where where i is an index for the message, T i a period of the message is and C i a transmission time for each transmission of the message, where the outbound port wait time w i,j The message is determined based on the fact that the median value calculated for the iteration is equal to the output port wait time w. i,jcalculated initial value for the iteration, and where the initial value for the lower bound (LB) or for the upper bound (UB) is adjusted before the next iteration of the iterative process based on the fact that the median value calculated for the iteration is not equal to the median value for the output port wait time w i,i The calculated initial value for the iteration is... [4] Vehicle (100) according to claim 1, wherein the measure to avoid the total waiting time w k during transmission, if the time limit for the message is exceeded, including the discarding of another message that precedes the message in a queue at one of the one or more switches (120) configured to forward the message. [5] Vehicle (100) according to claim 1, wherein the measure to mitigate the fact that the total waiting time w kduring a transmission exceeds the time limit for the message, including reducing a service level of a service of the vehicle (100) provided by communication of the message. [6] Method (300) for in-vehicle communication in a vehicle (100), wherein the method comprises the following: Sending a message from a sending control unit (110-T) of the vehicle (100) to be received by a receiving control unit (110-R); Forwarding the message between the sending control unit (110-T) and the receiving control unit (110-R) using one or more switches (120), wherein the sending control unit (110-T) and each of the one or more switches (120) include an output port for transmitting the message; Performing (310) a limited time control analysis to determine a total waiting time w kduring the transmission of the message from the sending control unit (110-T) to the receiving control unit (110-R) as a sum of all output port wait times w i,j to determine at all output ports to which the message encounters, whereby performing the limited timing analysis includes performing an iterative process and determining a lower bound (LB), an upper bound (UB), and a median value; and Taking measures to avoid or mitigate the total waiting time w k during transmission a deadline for the message is exceeded. [7] Method (300) according to claim 6, further comprising the control by the receiving control unit (110-R) of an operation of the vehicle (100) based on the message, wherein the limited timing control analysis is performed by the sending control unit (110-T), the receiving control unit (110-R) or a third control unit. [8] Method (300) according to claim 6, further comprising determining (320) an initial value for the lower bound (LB) and for the upper bound (UB), calculating the median value at each iteration of the iterative process as follows: median=(LB+UB)2, calculating an initial value for the exit port wait time w i,j at a given j-th of the output ports at each iteration of the iterative process as follows: wi,j=α+∑⌈medianTi⌉Ci,where where i is an index for the message, T i a period of the message is and C i a transmission time for each transmission of the message, determining the outbound port wait time w i,j the message is based on the fact that the median value calculated for the iteration is equal to that for the output port wait time w i,jcalculated initial value for the iteration, and adjusting the initial value for the lower bound (LB) or upper bound (UB) before the next iteration of the iterative process based on the fact that the median value calculated for the iteration is not equal to the initial value for the output port wait time w i,i The calculated initial value for the iteration is... [9] Method (300) according to claim 6, wherein taking the measure to avoid the total waiting time w k during transmission, if the time limit for the message is exceeded, including the discarding of another message that precedes the message in a queue at one of the one or more switches (120) configured to forward the message. [10] Method (300) Claim 6, wherein taking the measure to mitigate the fact that the total waiting time w kduring a transmission exceeds the time limit for the message, including reducing a service level of a service of the vehicle (100) provided by communication of the message.

Images