Method for testing a vehicle-mounted radio broadcast receiving device

Abstract

A method for testing a vehicle-mounted radio broadcast receiving device. The method comprises: splitting (S1) a radio signal (RF_SIG) received at a vehicle-mounted antenna (300) into two RF streams; and at the tested radio broadcast receiving device (100), converting the first RF stream into a demodulated audio signal and sending it to a test and recording device (400); at the test and recording device (400): receiving the demodulated audio signal, determining a spectrum that varies with time, and inputting the spectrum into an audio anomaly detection artificial intelligence module (452); receiving a second RF stream, converting it into a data signal, and recording the data signal into a temporary storage memory (460); and, when the artificial intelligence module (452) outputs a positive detection of an audio anomaly, transferring data from the temporary storage memory (460) to a permanent storage memory (470), the transferred data corresponding to a time window including the detected audio anomaly.

Description

Technical Field

[0001] The present disclosure relates to the field of testing vehicle-mounted radio broadcast receiving devices. Background Art

[0002] Today, modern vehicles are equipped with an in-vehicle infotainment system that includes hardware components and software components that provide audio and / or video entertainment. Part of the in-vehicle infotainment system is a radio broadcast reception device. It can receive different types of radio transmissions, such as analog signals (e.g. FM or AM signals) and digital transmissions (e.g. DAB (Digital Audio Broadcasting), DAB+, etc.).

[0003] In the illustrative example, the on-board radio receiving device is tuned to receive a broadcast source via DAB transmission, and at a certain point, the vehicle leaves the DAB range, which causes the radio receiving device to automatically switch from DAB to FM, so as to receive the same broadcast source via FM transmission. In the vehicle, the switch from DAB to FM should not be noticed by the driver who is listening to the audio signal output by the speaker. More generally, the driver should not be aware that the radio receiving device changes the source of the received radio signal. The switch should be transparent to the driver. However, in fact, this switch is likely to produce sound errors (also referred to as chirps), which results in unpleasant sound effects for the driver. This audio phenomenon is undesirable, and may be dangerous if the driver is disturbed by the sound effect. Other situations may cause audio anomalies or sound errors, such as interference phenomena or damage to the audio signal.

[0004] The on-board radio broadcast receiver is designed not to produce audio anomalies (in other words: unwanted unpleasant sound effects). Audio anomalies can be mitigated and / or suppressed and / or avoided. For example, the radio broadcast receiver can include a chirp blanking system. Ideally, the radio broadcast receiver should not produce any audio anomalies.

[0005] In order to build a radio receiver, the manufacturer needs to test it in a vehicle under real conditions. The undesirable phenomenon that produces audio anomalies (undesirable sound effects) is random and very difficult to reproduce in experiments. It is necessary to perform a series of cumbersome in-car tests, including sending a vehicle equipped with a tested radio receiver and driven by a driver for many kilometers on the road to capture the moment when the driver can hear the unpleasant sound effects. This test operation has many disadvantages. The first disadvantage is that it is time-consuming and resource-intensive because the driver needs to spend a long time driving the vehicle to frequently obtain a small amount of audio anomaly detection. The second disadvantage is that the detection of audio anomalies is performed "by ear", which is not completely reliable. The third disadvantage is that when the firmware of the radio receiver is changed or updated, the test operation must be performed again.

[0006] Therefore, there is a need to facilitate testing operations and reduce the human resources required to perform testing operations. Summary of the invention

[0007] The present disclosure relates to a method for testing a radio broadcast receiving device in a vehicle and recording data related to abnormal audio behavior of the tested radio broadcast receiving device, the method comprising the following steps: receiving a radio signal carrying audio information from a radio broadcast source by a receiver antenna; splitting the received radio signal into a first radio signal stream and a second radio signal stream by a splitter, and sending the first radio signal stream to the tested radio broadcast receiving device, and transmitting the second radio signal stream to a tester and a recording device; at the tested radio broadcast receiving device, converting the first radio signal stream into a demodulated audio signal, and sending the demodulated audio signal to the tester and the recording device; recording device; at the testing and recording device: the control part receives the demodulated audio signal via the first input terminal, determines the spectrum of the demodulated audio signal over time, and inputs the determined spectrum into the audio anomaly detection artificial intelligence (AI) module; the radio receiver part receives the second radio signal stream via the second input terminal, converts the second radio signal stream into a data signal, and records the data signal into a temporary storage device; and when the audio anomaly detection AI module outputs a positive detection of an audio anomaly, the control part controls to transfer data from the temporary storage device to the permanent storage device, and the transferred data corresponds to a time window including the detected audio anomaly.

[0008] The method can be used in the field (in other words: when the vehicle is driving on a road and can receive one or more radio broadcasting stations) or on a test bench (in other words: under laboratory conditions, two or more radio broadcasting sources are operating on the test bench) and allows the automatic detection of abnormal behavior of the device under test by an AI module, such as a neural network, which is fed as input the time-varying spectrum of the demodulated audio signal output by the device under test and allows the data carried by the RF signal causing the detected event to be saved. The saved data can be used to reproduce the detected event at a later time, without reusing the vehicle, for example for regression testing operations.

[0009] The method may also include a step of false detection performed by the control part of the testing and recording device, including receiving additional information related to the received radio signal obtained from a source outside the broadcast radio receiving device under test, and verifying based on the additional information that the positive detection from the AI ​​module is not a false detection; and wherein the step of controlling the transfer of data from the temporary memory to the permanent memory is performed only when the audio anomaly is not a false detection.

[0010] False positive detection allows identifying situations when a positive detection of an abnormal audio behavior of the device under test output by the AI ​​module (a "true" flag at the output of the AI ​​module) actually corresponds to a normal behavior (because the strange sound effect detected by the AI ​​module is caused by other actions such as user actions (e.g. an increase in the audio volume or a switch of a radio broadcast source). When this situation is detected (in other words: when a false positive is identified), the data are not transferred from the temporary memory to the permanent memory, which allows avoiding the permanent storage of data that are useless for later reproduction of the situation of interest in a test operation.

[0011] In an embodiment, the method further comprises the steps of obtaining, by a control portion of the testing and recording device, a message from an on-board communication network of the vehicle, and determining that the positive detection of the audio abnormality is a false detection based on the obtained message.

[0012] Messages from the vehicle's onboard communication network may be recorded in a temporary storage memory and, in the event that an audio anomaly is detected (ie, not a false detection), the messages corresponding to the time window including the detected audio anomaly are transferred from the temporary storage memory to the permanent storage memory.

[0013] In an embodiment, the method also includes the following steps: the radio receiver part of the test and recording device inputs the second radio signal stream into the reference broadcast radio receiver, the reference broadcast radio receiver converts the second radio signal stream into a reference demodulated audio signal, and sends the reference demodulated audio signal to the control part, the control part of the test and recording device determines the time-varying spectrum of the reference demodulated audio signal, and inputs the spectrum of the reference demodulated audio signal into the AI ​​module.

[0014] In one embodiment, the method includes the step of converting, by a radio receiver portion of the testing and recording apparatus, the second radio signal stream into a baseband data signal that is recorded in a temporary memory.

[0015] It is better to save the baseband signal which requires less storage space. To later replay (or reproduce) an event that has been recorded in permanent storage, the recorded baseband signal can be taken and shifted to the original spectrum in order to reconstruct a signal that is almost identical to the original signal converted to baseband and recorded.

[0016] In an embodiment, the method further comprises the following steps: in a radio receiver portion of the testing and recording device, applying a second radio signal stream to a first software-defined radio receiver and a second software-defined radio receiver; and wherein the first software-defined radio receiver is tuned to receive the radio broadcast source using a first frequency and a first modulation, and the second software-defined radio receiver is tuned to receive the radio broadcast source using at least one of a second frequency different from the first frequency and a second modulation different from the first modulation.

[0017] A radio broadcast source can broadcast programs via different transmission methods using different RF signal sources. For example, a radio broadcast source can broadcast the same program via FM (frequency modulation) transmission and DAB (digital audio broadcasting) or DAB+ transmission. The test and recording device has two (or more) software-defined radio receivers to receive and process RF signals broadcast by different transmission modes in parallel. This enables recording of all signals from both signal sources when switching from one signal source to another.

[0018] Advantageously, when the second radio signal stream is applied to the first software defined radio receiver and the second software defined radio receiver, the first software defined radio receiver outputs a first baseband data signal, the second software defined radio receiver outputs a second baseband data signal, the first baseband data signal and the second baseband data signal are recorded in a temporary memory, and in the step of sending data, the first baseband data signal and the second baseband data signal are transferred from the temporary memory to the permanent memory.

[0019] The method may further include the step of generating a test scenario based on the data stored in the permanent memory, and storing the test scenario in a database for later use in another test operation.

[0020] The method may further comprise the step of using the test scenario in an operation of regression testing the radio broadcast receiving apparatus being tested or in an operation of testing another radio broadcast receiving apparatus.

[0021] The present disclosure also relates to a testing and recording device, which is used to test a radio broadcast receiving device in a vehicle when a radio signal carrying audio information is received from a radio broadcast source via a receiver antenna in the vehicle, and to record data related to abnormal behavior of the radio broadcast receiving device under test, including: a first input end, which is used to receive a demodulated audio signal from the radio broadcast receiving device under test, the demodulated audio signal corresponding to a first radio signal stream split from the received radio signal; a control part, which receives the demodulated audio signal; a second input end, which is used to receive a second radio signal stream split from the received radio signal; a radio receiver part, which receives the second radio signal stream; a temporary memory; a permanent storage memory; wherein the control part and the radio receiver part perform corresponding steps of the previously defined method.

[0022] In one embodiment, the radio receiver portion receives a second radio signal stream from an antenna via a second input terminal, converts the second radio signal stream into a data signal, and records the data signal into a temporary memory; and the control portion receives a demodulated audio signal via a first input terminal, determines a spectrum that varies over time based on the demodulated audio signal, inputs the determined spectrum into an artificial intelligence (AI) module for audio anomaly detection, and when the AI ​​module outputs a positive detection of an audio anomaly, controls data transfer from the temporary memory to the permanent memory, the transferred data corresponding to a time window including the detected audio anomaly.

[0023] In an embodiment, the control part also includes a false detection detector, which receives a message from an on-board communication network of the vehicle and determines that a positive detection of an audio abnormality is a false detection based on the obtained message, and wherein the control part controls the data transfer from the temporary storage to the permanent storage only when the audio abnormality is not a false detection.

[0024] Advantageously, the radio receiver part converts the second radio signal stream into a baseband data signal before recording the second radio signal stream into the temporary memory.

[0025] In an embodiment, the radio receiver portion includes: a first software-defined radio receiver, which is tuned to receive the radio broadcast source RBS using a first frequency and a first modulation; and a second software-defined radio receiver, which is tuned to receive the radio broadcast source RBS using at least one of a second frequency different from the first frequency and a second modulation different from the first modulation, wherein a second radio signal stream is applied to the first software-defined radio receiver and the second software-defined radio receiver.

[0026] The disclosure also relates to a system comprising a testing and recording device as previously defined and a splitter for splitting a radio signal received at a receiver antenna. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Other features, objects and advantages of the present disclosure will become more apparent from a reading of the detailed description of non-limiting embodiments made with reference to the accompanying drawings.

[0028] Figure 1 A system for testing a radio broadcast receiving device in a vehicle and recording data related to abnormal behavior is shown according to an embodiment.

[0029] Figure 2 is a flow chart of a method for testing a radio broadcast receiving device in a vehicle according to an embodiment. DETAILED DESCRIPTION

[0030] Figure 1 The illustrated testing and recording system may be used to test the radio broadcast receiving device 100 in the field or on a test bench.

[0031] For on-site testing and recording operations, the radio broadcast receiving device 100 is installed in a vehicle 200, typically while the vehicle 200 is moving and a radio signal (or RF signal) carrying audio information is being received by a receiver antenna 300 in the vehicle 200 from a radio broadcast source RBS such as a radio station. i When receiving, the radio broadcast receiving device 100 under test records data related to abnormal audio behavior. The radio broadcast receiving device 100 under test can receive audio from multiple radio broadcast sources RBS i A radio signal (in other words: RF signal) is received, where i=1, 2, . . . The abnormal audio behavior of the tested radio broadcast receiving device 100 corresponds to an unwanted sound effect (in other words, a noise hazard, also called chirp), which is produced by a loudspeaker or more generally by an electroacoustic transducer at its output. The data recorded by the test and recording system can then be used by the scenario generator 700 to generate test scenarios, each scenario reproducing a situation related to the abnormal audio behavior (previously detected) of the tested radio broadcast receiving device 100. The test scenarios can be stored in the scenario database 701. The test scenarios can be used and replayed later in a regression test operation of the radio broadcast receiving device 100 (for example after a software change in the tested radio broadcast receiving device 100, and in a test operation of another radio broadcast receiving device).

[0032] The test and recording system has two main components: a test and recording device 400; and a splitter 500 for splitting the radio signal received at the receiver antenna 300.

[0033] The receiver antenna 300 is connected to the radio broadcast receiving device 100 under test and the test and recording device 400 via a splitter 500. In an embodiment, the splitter 500 is a passive device, such as a directional coupler. Alternatively, the splitter can be any other type of RF signal splitter. The splitter 500 is responsible for splitting the radio signal RF_SIG received at the receiver antenna 300 into a first radio signal stream RF_SIG1 and a second radio signal stream RF_SIG2, and sending the first radio signal stream RF_SIG1 to the radio broadcast receiving device 100 under test, and sending the second radio signal stream RF_SIG2 to the test and recording device 400.

[0034] The testing and recording device 400 is responsible for testing the radio broadcast receiving device 100 in the vehicle 200 while receiving a radio broadcast source RBS via the receiver antenna 300 in the vehicle 200. i The radio signal RF_SIG carrying audio information is received, any abnormal behavior causing unwanted sound effects in the tested radio broadcast receiving device 100 is detected, and data related to the detected abnormal behavior is recorded.

[0035] The testing and recording device 400 has: a first input terminal 410, which is used to receive a demodulated audio signal Audio_SIG, that is, an analog audio signal, from the radio broadcast receiving device 100 under test, which audio signal corresponds to a first radio signal stream RF_SIG1 split from the received radio signal RF_SIG by the splitter 500; a second input terminal 420, which is used to receive a second radio signal stream RF_SIG2 from the splitter 500.

[0036] At the output of the radio broadcast receiving apparatus 100, the demodulated audio signal Audio_SIG is an analog audio signal suitable for being sent to an electroacoustic transducer like a loudspeaker. It may comprise two audio signal parts for two left and right outputs on the transducer.

[0037] Optionally, the test and recording device 400 has a third input 430 and a message acquisition and interpretation 455 for receiving messages from the vehicle's onboard (in-vehicle) communication network of the vehicle 200. In an embodiment, the message is a CAN message from the vehicle's CAN (Controller Area Network) bus, through which the onboard microcontroller and multiple devices communicate with each other. The message sender 600 in the vehicle 200 is responsible for filtering all infotainment-related HMI messages (such as volume, radio button clicks, etc.) and mileage data (GPS, speed, yaw rate, etc.) exchanged via the communication network (CAN bus) in the vehicle, and sending the filtered messages and data to the test and recording device 400. The message sender 600 may include a software component running on the processor 480 of the test and recording device 400.

[0038] The CAN messages received by the testing and recording device 400 include, if any, relevant information (the following list is an illustrative and non-limiting example): user commands entered by a user via an HMI (human-machine interface) in the vehicle 200, in particular user commands for controlling the radio broadcast receiving device 100 under test (for example for increasing the audio volume on the speaker output or for switching radio broadcast sources and / or audio sources); location information of the vehicle 200, including, for example, the GPS location coordinates of the vehicle 200; movement information of the vehicle 200, including speed and optionally acceleration, heading and / or expected trajectory for at least x seconds in the future; information related to the radio signal RF_SIG received by the receiver antenna 300, including, for example, the strength, frequency and / or modulation of the received radio signal RF_SIG.

[0039] The testing and recording device 400 has a radio receiver portion 440 , a control portion 450 , a temporary memory 460 , and a permanent memory 470 .

[0040] In an embodiment, the radio receiver portion 440 includes a first software defined radio component 441a, a second software defined radio component 441b, and a reference radio broadcast receiving device 443 that is different from the radio broadcast receiving device 100 under test. The radio receiver portion 440 is connected to the second input terminal 420. The first software defined radio component 441a, the second software defined radio component 441b, and the reference radio broadcast receiving device 443 receive the first radio signal stream RF_SIG1 via the second input terminal 420.

[0041] The test and recording device 400 has at least two software defined radio receivers 441a, 441b, because a radio broadcast source may broadcast its programs via two or more transmission modes (e.g., FM, DAB, etc.). When the radio broadcast receiving device 100 receives a radio broadcast signal from a given radio broadcast source RBS via the antenna 300,i When receiving an RF signal having a given frequency called the first frequency f1 (e.g., in the case of FM transmission, a frequency within the FM band of 87.5-108 MHz) and a given modulation called the first modulation mod1 (e.g., a frequency modulation), the first software defined radio receiver 441a and the second software defined radio receiver 441b are programmed to automatically tune as follows: the first software defined radio receiver 441a is tuned to receive the radio broadcast source RBS using the first frequency f1 and the first modulation mod1 i The second defined radio receiver 441b is tuned to receive the same radio broadcast source RBS i , but using at least one of a second frequency f2 different from the first frequency f1 (eg a frequency in the DAB band of 174-240 MHz) and a second modulation mod2 different from the first modulation mod1 (eg OFDM or PSK modulation).

[0042] The frequency f1 and the modulation mod1 of the second radio signal stream received via the second input 420 are automatically determined by the software defined radio receiver.

[0043] For example, the first software defined radio receiver 441a is tuned to receive a given radio broadcast source RBS via FM transmissions. i , while the second software defined radio receiver 441b is tuned to receive the same radio broadcast source RBS via DAB (or DAB+) transmission i .

[0044] The reference radio broadcast receiving device 443 is different from the tested radio broadcast receiving device 100. For example, the two devices are manufactured by two different manufacturers or are different products manufactured by the same manufacturer. In operation, the second radio signal stream RF_SIG2 is applied to the reference radio broadcast receiving device 44 at the input.

[0045] In other embodiments, the radio receiver section 440 has more than two software defined radio receivers. One of these software defined radio receivers is automatically tuned to receive the radio broadcast source RBS using a first frequency f1 and a first modulation mod1 (i.e., the same frequency and modulation as set in the radio broadcast receiving apparatus 100 under test). i , and each other software defined radio receiver is tuned to receive the same radio broadcast source RBS i , but using a frequency and / or a modulation that is different from the frequency and modulation set in the radio broadcast receiving device 100 under test.

[0046] The software defined radio receivers 441a, 441b are responsible for converting the second radio signal stream RF_SIG2 into a data signal to be recorded in the temporary memory 460. In an embodiment, the radio signal stream RF_SIG2 is converted into a baseband data signal by each radio defined software receiver 441a, 441b. The software defined radio receivers 441a, 441b are tuned to acquire one or more parts of the spectrum of the received radio signal RF_SIG2 and move them to the baseband frequency. Each radio defined receiver 441a, 441b converts the input signal from the original frequency to the baseband frequency.

[0047] The data stream at the output of the radio-defined software receiver 441a, 441b is continuously saved in a temporary memory 460. The temporary memory 460 is, for example, a buffer with a given storage capacity, such as a circular buffer. For example, it can record the equivalent of at least a 40-second time window of the received radio signal RF, because the recorded data should cover a minimum of 20 seconds before and after an event such as an audio anomaly. In this way, the detected event can be reproduced at a later time under laboratory conditions. The advantage of saving the baseband signal is that it does not require too much space in the memory.

[0048] The control portion 450 is connected to the first input terminal 410 and the third input terminal 430 , and receives the demodulated audio signal Audio_SIG corresponding to the first radio signal stream RF_SIG1 from the tested radio broadcast receiving apparatus 100 via the first input terminal 410 , and receives a CAN message from the vehicle 200 via the third input terminal.

[0049] The control part 450 has a signal analyzer 451, an audio anomaly detection artificial intelligence (AI) module 452, a CAN message acquisition and interpreter 455, a false detection detector 453, and a generator 454 that generates a control signal for transferring data from the first temporary memory 460 to the second permanent memory 470.

[0050] A central unit or processor 480 is also provided for controlling the operation of the test and recording apparatus 400. In an embodiment, the signal analyzer 451, the AI ​​module 452, the false detection detector 453, the CAN message acquisition and interpreter 455 and the generator 454 are software components running on the processor 480.

[0051] The signal analyzer 451 is responsible for determining the time-varying spectrum (i.e., frequency variation over time) of the demodulated audio signal Audio_SIG. The signal analyzer 451 may use a fast Fourier transform. The audio anomaly detection artificial intelligence (AI) module 452 is, for example, a machine learning module (such as a neural network) that is trained to detect unwanted sound effects in the analog audio signal AUDIO_SIG.

[0052] The neural network 452 is responsible for detecting abnormal behavior of the radio broadcast receiving device 100 that causes unwanted acoustic effects on the speaker output. In an embodiment, this detection is performed based on the spectrum of the audio demodulated signal (analog) at the output of the radio broadcast receiving device 100 over time.

[0053] The training operation of the neural network 452 according to an embodiment will now be described. During training, the neural network 452 is fed with training data, which includes a time-varying frequency spectrum of an audio sample as input data, and a true (positive detection of an audio anomaly) or false (no audio anomaly detected) flag as output data.

[0054] In the first stage of the training operation, the audio samples used as training data are of different types. For example, they include music, advertisements, and speech. The input training data is preferably perfect data that is not distorted (in other words: no chirping, no noise, and no some strange signals). Many different styles of music are used for training. The different types of input training data (music, advertisements, and speech) fed to the neural network 452 during the first stage of training should not generate an alarm at the output of the neural network 452. In other words, in the first stage of training, the neural network 452 should output a pseudo (distortion-free) mark, which means that everything is fine, and does not output any positive detection of the unwanted sound effects corresponding to the abnormal audio behavior of the radio broadcast receiving device being tested.

[0055] In the second phase of the training operation, distorted (in other words: corrupted) data is used for training. Distorted data also includes samples of music, advertisements and speech. However, in this case, the training data includes audio anomalies or distortions, such as chirping or any other unwanted sound effects. The distorted data can be recorded data (e.g., data recorded during work to collect data using a test vehicle with a broadcast radio receiver), or data generated by generating distortions (e.g., by randomly cutting the data and switching the order of the cut data, and / or by adding some signals to the data to generate distortions, and / or by programmatically modifying the intensity of the audio signal, etc.). The distorted data (more precisely, the spectrum of the distorted data) is expected to generate an alarm (in other words: a true flag), corresponding to a positive detection of an audio anomaly (unwanted sound effect) at the output of the neural network 452. In training, the true flag is set to the output of the neural network 452 only when audio distortion occurs.

[0056] The process of labeling the training data is manual. In other words, the training uses data labeled by humans. For example, a log of x minutes of distorted data (e.g., 60 minutes) is output by a speaker, and a person with a mouse, for example, places a marker on the data where a problem (distortion) can be heard. In other words, the person listens to the x-minute log of distorted data and indicates the time points or time periods when the marker is true (in other words: when the audio problem can be heard). The distorted data is manually labeled or "ground-truthed". In this way, the neural network 452 is trained to detect distortions that sound unpleasant. The function of the neural network 452 is to detect "problems", and problems mean sounds that sound unpleasant.

[0057] The false detection detector 453 is responsible for detecting that a positive detection by the neural network 452 is actually a false detection. In operation, the false detection detector 453 receives additional information related to the received radio signal RF_SIG from an information source located outside the radio broadcast receiving device 100 under test, and in the case of a true flag at the output of the neural network 452, verifies based on the additional information that it is not a false detection.

[0058] In an embodiment, the additional information comprises at least one of the following two information elements: information of a CAN message received from the vehicle 200 via the third input 430 ; a time-varying spectrum of the demodulated audio signal at the output of the reference radio broadcast receiving device 443 .

[0059] As will be explained later in the description of the method, the false detection detector 453 uses the CAN message from the vehicle 200 and the demodulated audio signal at the output of the reference radio broadcast receiving device 443 to determine whether a positive detection (true flag) at the output of the neural network 452 is a false detection. In one embodiment, the output of the detector 453 is a signal indicating when a false detection has been identified (detected).

[0060] In one embodiment, the generator 454 receives at input the output of the neural network 452 and the output of the false detection detector 453. The generator 454 generates a control signal for transferring data from the temporary storage 460 to the permanent storage 470 only when a true flag is received from the neural network 452 and the signal at the output of the detector 453 does not indicate a false detection.

[0061] The false detection detector 453 is described as a different component from the AI ​​module 452. However, the functionality of false detection detection can be integrated into the AI ​​module 452. In this case, in addition to the time-varying spectrum of the demodulated audio signal at the output of the tested radio broadcast receiving device 100, the AI ​​module 452 receives as input the CAN message from the vehicle 200 and the time-varying spectrum of the demodulated audio signal at the output of the reference radio broadcast receiving device 443. In operation, the AI ​​module 452 can detect suspicious behavior of the tested radio broadcast receiving device 100 based on the time-varying spectrum of the demodulated audio signal at the output of the tested radio broadcast receiving device 100, and detect false detections using the CAN message and the audio signal from the output of the reference radio broadcast receiving device 443. The AI ​​module 452 is trained to be trained using input training data, which also includes the CAN message and the spectrum of the audio signal from the output of the reference radio receiver (in addition to the audio samples).

[0062] The signal analyzer 451, the neural network 452, the false detection detector 453, the generator 454 and the software defined radio receivers 441a, 441b may be software components stored in one or more memories and run on the processor 480. Therefore, the present disclosure also relates to a computer program, which includes instructions for causing the test and recording apparatus 400 to perform the steps (i.e., steps S6, S7, S8, S10, S11, S12, S20, S21 and S30) performed by the elements of the test and recording apparatus 400, and the method will now be described. The program instructions are stored in a storage module and can be executed by a computer.

[0063] refer to Figure 2, a method for testing a radio broadcast receiving device 100 under test in a vehicle 200 and recording data related to any abnormal behavior of the radio broadcast receiving device 100 under test that causes unwanted sound effects according to an embodiment will now be described. Assume that the vehicle 200 driven by a driver is moving and the radio broadcast receiving device 100 is tuned to receive a radio broadcast source RBS i As an illustrative and non-limiting example, a radio broadcast source RBS i There are two radio broadcast sources: a first source broadcasting FM radio signals using a first frequency f1 (within the portion of the radio spectrum from 87.5 MHz to 108 MHz) and a first modulation mod1 (FM modulation); a second source broadcasting DAB radio signals using a second frequency f2 (within the portion of the radio spectrum from 174 MHz to 240 MHz or from 1452 GHz to 1492 GHz) and a second modulation mod2 (e.g. OFDM).

[0064] Any other type of radio broadcast method may be used in the present disclosure, such as HD Radio or any satellite radio (eg, SiriusXM).

[0065] In a first step S0, a radio broadcast source RBS is received by a receiver antenna 300. i A radio signal RF_SIG carrying audio information is received. For example, the radio signal RF_SIG has a first frequency f1 within the portion of the radio spectrum from 87.5 MHz to 108 MHz and uses frequency modulation (FM) mod1.

[0066] In a second step S1, the received radio signal RF_SIG is split into a first radio signal stream RF_SIG1 and a second radio signal stream RF_SIG2 by the splitter 500. In step S2, at the output of the splitter 500, the first radio signal stream RF_SIG1 is sent to the radio broadcast receiving device 100 under test, and the second radio signal stream is sent to the testing and recording device 400.

[0067] In step S3, the first radio signal stream RF_SIG1 is received at the tested radio broadcast receiving device 100 and processed in step S4. In step S4, the tested radio broadcast receiving device 100 converts the first radio signal stream RF_SIG1 into a demodulated audio signal Audio_SIG. More precisely, the tested radio broadcast receiving device 100 uses a filter to separate the desired radio frequency signal from all other signals picked up by the antenna 300, increases the power of the signal for further processing by an amplifier, and finally recovers the information carried by the received radio signal RF_SIG1 by demodulation. The tested radio broadcast receiving device 100 outputs an audio demodulated signal (analog signal) Audio_SIG suitable for being input to an electroacoustic transducer (such as a speaker).

[0068] In step S5, the audio demodulated signal Audio_SIG at the output of the tested radio broadcast receiving apparatus 100 is sent to the first input terminal 410 of the testing and recording apparatus 400. At the testing and recording apparatus 400, the control section 450 receives the demodulated audio signal via the first input terminal 410.

[0069] In step S6, the signal analyzer 451 determines the time-varying frequency spectrum of the demodulated audio signal Audio_SIG received via the first input terminal 410. The frequency spectrum of the demodulated audio signal Audio_SIG is input to the neural network 452.

[0070] In step S10, in parallel with step S3, the radio receiver portion 440 of the test and recording device 400 receives the second radio signal stream RF_SIG2 via the second input terminal 420. Then, in step S11, each of the radio-defined software receivers 441a, 441b processes the received radio signal RF_SIG2 by filtering the desired radio frequency signal corresponding to its tuning frequency (f1 for the receiver 441a and f2 for the receiver 441b) and by converting the filtered radio signal into baseband data signals data_SIGa and data_SIGb. Then, in step S12, each of the radio-defined software receivers 441a, 441b outputs its baseband data signals data_SIGa and data_SIGb, which are recorded in the temporary memory 460. In one embodiment, the memory 460 is a circular buffer. Therefore, as soon as the buffer 460 is full, writing in the buffer 460 is performed as follows: when new data is added in the buffer 460, it overwrites the oldest data.

[0071] In step S20, which is parallel to steps S3 and S10, the message acquisition and interpreter 455 receives a CAN message from the vehicle 200. In step S21, the received CAN message is recorded in the temporary memory 460.

[0072] In step S7, the neural network 452 performs detection of audio anomalies in the demodulated audio signal Audio_SIG corresponding to abnormal behavior of the tested radio broadcast receiving apparatus 100. When abnormal behavior of the tested radio broadcast receiving apparatus 100 that causes undesirable sound effects is detected, a true flag is generated at the output of the neural network 452. A true flag is generated only when an audio anomaly is detected. When no abnormal audio behavior is detected, the neural network 452 outputs a false flag.

[0073] In one embodiment, the time-varying frequency spectrum of the reference audio demodulated signal at the output of the reference radio broadcast receiving device 443 may also be analyzed by the AI ​​module 452 . There are four possible scenarios: when the audio signal from the tested radio broadcast receiving device 100 and the audio signal from the reference radio broadcast receiving device 443 are both cleared (no anomaly is detected by the neural network 452), no true flag is generated at the output of the neural network 452; when the audio signal from the tested radio broadcast receiving device 100 generates a positive detection of an audio anomaly but not the audio signal from the reference radio broadcast receiving device 443, a true flag is generated at the output of the neural network 452; when the audio signal from the tested radio broadcast receiving device 100 is cleared (no audio anomaly is detected) and the audio signal from the reference radio broadcast receiving device 443 generates a positive detection of an audio anomaly, no true flag is generated at the output of the neural network 452; when the audio signal from the tested radio broadcast receiving device 100 and the audio signal from the reference radio broadcast receiving device 443 both generate positive detection of an audio anomaly, a true flag is generated at the output of the neural network 452 (this double positive detection may be due to poor radio signal strength, but this situation is worth recording and should be manually verified).

[0074] In the case where the neural network 452 outputs a true flag, in step S8, the false detection detector 453 performs a false detection. To this end, the detector 453 receives additional information related to the received radio signal RF_SIG, which is obtained from a source external to the tested radio broadcast receiving device 100, and verifies based on the additional information that the positive detection (true flag) from the neural network 452 is not a false detection. In one embodiment, the additional information includes a CAN message received from the vehicle via the third input terminal 430.

[0075] The detector 453 receives CAN messages containing information related to user commands entered by the user via the vehicle's human-machine interface, such as a command to increase the volume of the audio output by the speakers or to switch the audio source (e.g. from the current radio broadcast source RBS to the current radio broadcast source RBS). i Switch to another radio broadcast source RBS j ) command. In the event that the neural network 452 outputs a positive detection of an audio anomaly, the detector 453 verifies whether the positive detection is a false detection based on the obtained CAN message. For example, if information about an increase in volume or a switch in the radio broadcast source is provided by the CAN message, the detector 453 determines that the positive detection (true flag) from the neural network 452 is a false detection because the sound effect detected by the neural network corresponds to the normal behavior of the radio broadcast receiving device 100 being tested. Typically, if the CAN data indicates that the user has changed some settings that will affect the volume, the flag from the AI ​​module 452 can be ignored. The behavior in this case should be user-configurable.

[0076] When the CAN message does not provide any information about commands or signals from the on-board HMI that may have caused an audible effect that is likely similar to audio distortion but is actually normal, the detector 453 determines that the positive detection (true flag) of the audio anomaly from the neural network 452 is not a false positive.

[0077] In the case where no false detection is detected in step S8 , the generator 454 generates a control signal for transferring data from the first temporary memory 460 to the second permanent memory 470 .

[0078] As previously described, false positive detection may be performed directly by the neural network 452 .

[0079] In step S30 , data is transferred from the first temporary storage 460 to the second permanent storage 470 .

[0080] The transferred data corresponds to a time window including the detected audio anomaly. It covers at least N seconds (usually for a very short period of time) before and after the detected audio anomaly. For example, N can be 15 or 20 seconds or more. The transferred data includes: data from the baseband data signal data_SIGa of the software receiver 441a of the first radio design; data from the baseband data signal data_SIGb of the software receiver 441b of the second radio design; CAN messages from the vehicle 200 and received via the third input terminal 430.

[0081] Optionally, the transferred data may further include: data corresponding to clearing the buffer 460 into the permanent memory 470 to save the entire baseband data signal; values ​​of internal data from the vehicle (RMS value, radio signal strength, GPS location, etc.); and the current state of the FSM (completion state machine) of the radio broadcast receiving device 100 being tested.

[0082] The method can be used under laboratory conditions on a test bench where two or more radio broadcast sources are operated.

[0083] The data recorded in the permanent memory 470 can be used to reproduce the detected events that caused the abnormal behavior of the tested radio broadcast receiving device 100 under laboratory conditions without reusing the vehicle driven by the driver. The test scenario can be generated by the scenario generator 700 based on the data stored in the permanent memory 470. The generated test scenario is stored in the scenario database 701 for example for later use in another test operation. The test scenario can be used in the operation of regression testing of the tested radio broadcast receiving device 100 or in the operation of testing another radio broadcast receiving device.

[0084] The present disclosure provides many advantages: the use of the test and recording device 400 enables the construction of a RWUP (Real World User Profile) database (a recorded catalog of many situations in which problems occur with a radio receiver), which can be successfully used as a basic source for regression testing; vehicle rides do not have to be performed by experienced engineers, because the test and recording device 400 automatically collects only relevant fragments of the test drive and stores metadata (such as fault category, vehicle speed, GPS data, etc.), thereby enabling initial data filtering; each detected anomaly can be reproduced at any time later and used for regression testing and new tests; it allows reducing the time spent in testing and driving costs.

Claims

1. A method for testing a radio broadcast receiving device (100) in a vehicle (200), the method comprising the following steps: The receiver antenna (300) receives the radio broadcast source (RBS) i ) receiving (S0) a radio signal (RF_SIG) carrying audio information; Splitting (S1) a received radio signal (RF_SIG) into a first radio signal stream (RF_SIG1) and a second radio signal stream (RF_SIG2) by a splitter (500), sending the first radio signal stream (RF_SIG1) to a radio broadcast receiving device (100) under test, and sending the second radio signal stream (RF_SIG2) to a testing and recording device (400); At the radio broadcast receiving device (100) under test, converting (S4) the first radio signal stream (RF_SIG1) into a demodulated audio signal (Audio_SIG), and sending (S5) the demodulated audio signal (Audio_SIG) to the testing and recording device (400); At the testing and recording device (400): The control part (450) receives the demodulated audio signal via the first input terminal (410), determines (S6) a spectrum of the demodulated audio signal (Audio_SIG) that varies with time, and inputs (S7) the determined spectrum into an audio anomaly detection artificial intelligence (AI) module (452); receiving (S10) the second radio signal stream (RF_SIG2) via the second input terminal (420) by a radio receiver section (440), converting (S11) the second radio signal stream into baseband signal data (Data_SIGa, Data_SIGb), and recording the baseband signal data into a temporary storage memory (460); as well as When the audio anomaly detection AI module (452) outputs a positive detection of an audio anomaly, the control part (450) controls (S30) the transfer of baseband signal data from the temporary storage memory (460) to the permanent storage memory (470) for reproducing the event causing the detected anomaly, the transferred baseband signal data corresponding to a time window including the detected audio anomaly.

2. The method according to claim 1, further comprising the following steps: A message is obtained (S20) by the control portion (450) of the testing and recording device (400) from an on-board communication network of the vehicle (200), and it is determined (S8) based on the obtained message that the positive detection of the audio abnormality is a false detection.

3. The method according to claim 2, wherein: The message from the on-board communication network of the vehicle (200) is recorded (S21) in the temporary storage memory (460), and when an audio anomaly that is not a false detection is detected, the message corresponding to the time window including the detected audio anomaly is transferred from the temporary storage memory to the permanent storage memory.

4. The method according to claim 1, comprising the steps of: The radio receiver section of the test and recording device inputs the second radio signal stream (RF_SIG2) to a reference radio broadcast receiving device (443), the reference radio broadcast receiving device (443) converts the second radio signal stream (RF_SIG2) into a reference demodulated audio signal, and sends the reference demodulated audio signal to the control section (450); as well as The control portion (450) of the test and recording device (400) determines a time-varying frequency spectrum of the reference demodulated audio signal, and inputs the frequency spectrum of the reference demodulated audio signal to the audio anomaly detection AI module (452).

5. The method according to claim 1, further comprising the steps of: applying the second radio signal stream (RF_SIG2) to a first software defined radio receiver (441a) and a second software defined radio receiver (441b) at the radio receiver portion (440) of the test and recording apparatus (400), and wherein the first software defined radio receiver (441a) is tuned to receive the radio broadcast source (RBS) using a first frequency (f1) and a first modulation (mod1) i ), and the second software defined radio receiver (441b) is tuned to receive the radio broadcast source using at least one of a second frequency (f2) different from the first frequency (f1) and a second modulation (mod2) different from the first modulation (mod1).

6. The method according to claim 5, wherein: When the second radio signal stream (RF_SIG2) is applied to the first software defined radio receiver (441a) and the second software defined radio receiver (441b), the first software defined radio receiver (441a) outputs first baseband signal data (Data_SIGa), and the second software defined radio receiver (441b) outputs second baseband signal data (Data_SIGb), the first baseband signal data (Data_SIGa) and the second baseband signal data (Data_SIGb) are recorded in the temporary storage memory (460), and, in the step of transferring (S30) data, the first baseband signal data (Data_SIGa) and the second baseband signal data (Data_SIGb) are transferred from the temporary storage memory (460) to the permanent storage memory (470).

7. The method according to claim 1, further comprising the steps of: A test scenario is generated based on the data stored in the permanent storage memory, and the test scenario is stored in a database for later use in another testing operation.

8. The method according to claim 7, further comprising the steps of: The test scenario is used in an operation of a regression test of the tested radio broadcast receiving device or in an operation of testing another radio broadcast receiving device.

9. A test and recording device (400) for detecting a radio broadcast source (RBS) received by a receiver antenna (300) in a vehicle (200). i ) while receiving a radio signal carrying audio information, testing a radio broadcast receiving device (100) in the vehicle (200), the testing and recording device (400) comprising: a first input terminal (410) receiving a demodulated audio signal (Audio_SIG) from the radio broadcast receiving device (100) under test, the demodulated audio signal corresponding to a first radio signal stream (RF_SIG1) split from a received radio signal (RF_SIG); a control section (450) receiving the demodulated audio signal (Audio_SIG); a second input terminal (420) receiving a second radio signal stream (RF_SIG2) split from the received radio signal (RF_SIG); a radio receiver portion (440) that receives the second radio signal stream (RF_SIG2); Temporary storage memory (460); as well as Persistent storage memory (470), Wherein, the control part (450) and the radio receiver part (440) perform corresponding steps of the method according to any one of claims 1 to 8.

10. The testing and recording device according to claim 9, wherein: The radio receiver portion (440) comprises: A first software defined radio receiver (441a) is tuned to receive the radio broadcast source (RBS) using a first frequency (f1) and a first modulation (mod1). i );as well as a second software defined radio receiver (441b) tuned to receive the radio broadcast source (RBS) using at least one of a second frequency (f2) different from the first frequency (f1) and a second modulation (mod2) different from the first modulation (mod1) i ), The second radio signal stream is applied to the first software defined radio receiver (441a) and the second software defined radio receiver (441b).

11. A system comprising a testing and recording device (400) according to claim 9 or 10 and a splitter (500) for splitting a radio signal (RF_SIG) received at the receiver antenna (300).

Images