Echo cancellation (AEC) enables comfortable full-duplex hands-free calls in the car. The Clarity CVC-HFK introduced in this article can provide an integrated single-amplifier solution (OMS) noise suppression algorithm.
It supports adaptive noise cancellation, which can reduce the environmental noise in the microphone (incoming) signal, extract the required speech, and transmit clear speech (outgoing) to the remote user.
Two factors determine the necessity of using a hands-free cellular phone system. The first is that the terminals of current cellular phone systems are mostly handheld, which causes inconvenience to drivers. Drivers often have to put down their phones and drive with two hands, such as turning, etc., and then turning back to talk. The interruption of telephone conversations is inconvenient and may even cause economic cost losses because mobile telecommunications charges are quite high. Another important factor is safety. Imagine a driver who uses only one hand to control the steering wheel and is on the phone. Since the driver can't control the steering wheel with both hands, what anti-lock braking system and airbags would be meaningless. Therefore, the hands-free cellular phone system is becoming a necessity for drivers using mobile phones.
Figure 1: The structure of the hands-free kit development platform HFK development platform
The HFK development platform is a set of solutions including DSP, providing software and hardware design, which can realize the rapid development of the final product and make it unique. The HFK development platform can be connected to the TI Software Development Environment Code Composer StudioTM (CCStudio) development tool through JTAG. The combination of the development environment and documentation enables the rapid integration of TI DSP third-party software and accelerates the time-to-market of products. The HFK development platform is suitable for high-quality automotive aftermarket hands-free kits that require high cost and performance, as well as Bluetooth-enabled HFK. Echo cancellation software for system-free
As far as the car hands-free radio / telephone system is concerned, one disadvantage is that the far-end speaker will feel an echo. To create a comfortable full-duplex hands-free conversation environment in the car, the most important software link is the echo canceller (AEC). The European Telecommunications Standards Institute (ETSI) is currently developing standards for the AEC system.
The echo phenomenon is caused by the coupling between the speaker and the microphone. In full-duplex communication, the far-end speaker will have a delay in hearing its own voice. The length of the delay depends on the delay inside the car and the Global System for Mobile Communications (GSM). The echo path length is a key parameter of AEC.
Figure 2: CVC-HFK application diagram
Adaptive filtering (more accurately, the NLMS algorithm) is one of the most common solutions for AEC. The NLMS algorithm achieves a good compromise between the amount of calculation and performance.
The other issue related to AEC is the fuzzy tone (DT) situation when the two are talking at the same time. If it is not detected, DT will cause the divergence of the adaptive algorithm.
The AEC software uses the NLMS algorithm to cancel the echo, which is executed by the C54x DSP assembler.
1. NLMS algorithm
The NLMS algorithm can update the coefficients of the adaptive finite impulse response (FIR) filter, which is used to predict the echo, and then we subtract the predicted value from the actual echo to give the residual echo.
2. Active channel detection
A key feature of the AEC algorithm is active channel detection. When the far-end operator is silent and the near-end operator speaks, because the near-end operator is no longer an echo, the filter does not need to be adapted. By calculating the signal energy and comparing this energy with the adaptive threshold, it can be achieved Detection of active channels.
3. Detection of fuzzy tone (DT)
In the case of DT, the near-end signal on the loudspeaker includes echo and near-end speech (ie, vague tone). The residual error used to update the filter coefficients includes near-end speech, and if the algorithm is still adapting, the algorithm may begin to diverge, and this must be avoided. DT detection uses an energy-based algorithm and a variable threshold to solve this problem.
4. Benchmark
The benchmark of the AEC software (expressed in 16-bit words) is:
Code size: 154 words;
Static RAM: 527 words;
Erase RAM: 2 words;
The maximum calculation cost is 4.7 MIPS. The computational consumption is the largest in the ST period, and drops to 2.4 MIPS in the DT period. The ST period occupies the main part of the call, while the DT period only appears in shorter individual cases.
CVC-HFK software
CVC-HFK (Clear Voice Capture-Hands Free Kit) integrates functions such as echo cancellation, noise suppression, and nonlinear processing, and is an optimized HFK solution. The CVC-HFK solution uses a comprehensive adaptive sub-band method to improve the performance of the main aspects while taking up very few resources. In the automotive environment, environmental noise is the main problem to be overcome by hands-free systems. Therefore, in addition to the echo cancellation function, Clarity CVC-HFK also provides an integrated single microphone solution (OMS) noise suppression algorithm. The OMS solution supports adaptive noise cancellation, which can attenuate the environmental noise in the microphone signal (incoming), extract the required speech, and transmit clean speech (outgoing) to the far-end user. Since CVC-HFK is fully adaptive, there is no need to adjust too much. Below, we will introduce the CVC-HFK solution and its main aspects of performance.
1. CVC-HFK AEC
The CVC-HFK echo canceller is a "stateless" AEC, which uses a variant of the standard frequency domain NLMS algorithm as its main adaptive filter. We will explain the benefits of using these methods below. First, the sub-band frequency domain method can uncorrelate or whiten the input signal in each band, which can achieve faster convergence compared to AEC in a considerable time domain. Second, stateless AEC enables continuous filter adaptation, which can improve robustness and overall fuzzy sound performance in noisy environments. As mentioned earlier, in the case of DT, the microphone signal contains both echo and near-end speech. The near-end voice is not associated with the echo signal, and if there is no process to avoid it, it will cause the divergence of the adaptive filter. Third, NLMS can achieve uniform convergence independent of input amplitude.
Because of this, CVC-HFK AEC can obtain a typical 40dB ERLE (echo return loss gain value), up to 50dB ERLE, and can achieve a fast convergence time of about 80ms, and can perform full-duplex operation in most environments . In addition, CVC-HFK AEC uses a tail length of 64ms for its adaptive filter, which allows greater flexibility in terms of internal capacity.
2. CVC-HFK NS (Noise Suppressor)
The CVC-HFK noise suppressor is a frequency domain algorithm that uses voice and noise characteristics to help extract voice from synthesized noise and voice signals. The two main modules of CVC-HFK NS are speech composition analysis and speech extraction.
The speech component analysis module uses the temporal and related attributes of speech and noise to construct a predictable model of speech composition. The speech extraction block can modify each frequency component according to the speech and noise model. In addition, the speech extraction block can also make full use of the principle of sound quality to minimize noise floor and perceived speech distortion.
CVC-HFK NS adopts this scheme to achieve 10-15dB SNR (signal-to-noise ratio) improvement in a noisy environment while maintaining good voice quality. In an extremely low noise environment where the SNR is already high enough, because the NS is turned off, no speech distortion will occur.
3. CVC-HFK NLP (non-linear processing)
As the system distortion increases, CVC-HFK NLP is minimal. The amount of distortion added by CVC-HFK NLP is much lower than standard NLP modules such as center clippers because it uses information from the input and error signals to determine additional attenuation. Since all CVC-HFK modules use frequency-domain algorithms, compared to solutions that use both time-domain and frequency-domain algorithms, they can significantly save memory and simplify computational complexity.
System integration design
When integrating a TI-HFK board with a cellular hands-free kit, several components and appropriate interfaces are required to achieve a good mobile call.
You must choose components that are compatible with both CVC-HFK application software and board hardware to achieve good performance. HFK can support various loudspeakers, loudspeakers and car audio systems. However, in order to reduce the changes to the application manual, we have selected dedicated industry standard components, which will greatly help your successful adjustment. Three connections are required from the TI-HFK board to the cellular kit to achieve integration: the loudspeaker inputs the TI HFK board; the outgoing, processed audio output; input, and the incoming signal received from the cellular kit.
Here are some suggestions for the design of loudspeaker and loudspeaker placement.
1. The position and orientation of the loudspeaker In order to achieve the best overall microphone performance, some key variables should be understood before the device is finally installed in the car. It is recommended to keep the distance between the microphone and the user's mouth in the car at 46cm (18 inches). The recommended distance ranges from 30 to 56cm (12-22 inches).
2. Avoid exposure of the loudspeaker to the airflow (windows and fans) as much as possible;
3. Appropriately consider the size and installation scheme of the loudspeaker so that the front of the loudspeaker can be aligned with the mouth of the user in the car. Based on the above considerations, with the help of Figure 3, you can choose the best loudspeaker position. First, follow the recommendations in Priority Zones 1, 2 and 3. Once a decision is made, you can fix the loudspeaker with a metal plate or velcro strap, and you can connect the cable back to the electronic device for termination. Afterwards, the concealment of the cable should be ensured to maintain its aesthetic appearance, and in addition, the cable should be kept fixed so as not to be pinched or knotted. Finally, avoid connecting parallel cables to antenna connectors and other noisy cables.
4. Speaker position It is recommended to install the speaker in an appropriate location to provide good voice performance without disturbing the pickup area of ​​the microphone. The pickup area is a +30 degree cone, which protrudes outward from the front of the microphone to the driver of the car.
Figure 3: Speaker placement
The speaker should be located at least 1 meter (3 feet) away from the microphone. The speaker should be far away from the pickup area of ​​the microphone to reduce the chance of echo feedback. Ideally, the speakers should be located behind the front of the microphone or at 90 degrees. The size of the speaker's sound distortion will have a direct negative impact on the echo.
Summary of this article
The use of hands-free systems continues to gain popularity, and users also expect that performance will continue to improve. Given the many options available for HFK implementation, it is obvious that integrating software algorithms and hardware signal processors is a thoughtful move, which will be very beneficial. The HFK development kit can cope with all the above problems and bring benefits to those who develop or sell the above products. The TI TMS320C5407 development kit with AEC and CVC-HFK provides the required flexibility and high performance, and can quickly and inexpensively bring HFK solutions to market.
It supports adaptive noise cancellation, which can reduce the environmental noise in the microphone (incoming) signal, extract the required speech, and transmit clear speech (outgoing) to the remote user.
Two factors determine the necessity of using a hands-free cellular phone system. The first is that the terminals of current cellular phone systems are mostly handheld, which causes inconvenience to drivers. Drivers often have to put down their phones and drive with two hands, such as turning, etc., and then turning back to talk. The interruption of telephone conversations is inconvenient and may even cause economic cost losses because mobile telecommunications charges are quite high. Another important factor is safety. Imagine a driver who uses only one hand to control the steering wheel and is on the phone. Since the driver can't control the steering wheel with both hands, what anti-lock braking system and airbags would be meaningless. Therefore, the hands-free cellular phone system is becoming a necessity for drivers using mobile phones.
Figure 1: The structure of the hands-free kit development platform HFK development platform
The HFK development platform is a set of solutions including DSP, providing software and hardware design, which can realize the rapid development of the final product and make it unique. The HFK development platform can be connected to the TI Software Development Environment Code Composer StudioTM (CCStudio) development tool through JTAG. The combination of the development environment and documentation enables the rapid integration of TI DSP third-party software and accelerates the time-to-market of products. The HFK development platform is suitable for high-quality automotive aftermarket hands-free kits that require high cost and performance, as well as Bluetooth-enabled HFK. Echo cancellation software for system-free
As far as the car hands-free radio / telephone system is concerned, one disadvantage is that the far-end speaker will feel an echo. To create a comfortable full-duplex hands-free conversation environment in the car, the most important software link is the echo canceller (AEC). The European Telecommunications Standards Institute (ETSI) is currently developing standards for the AEC system.
The echo phenomenon is caused by the coupling between the speaker and the microphone. In full-duplex communication, the far-end speaker will have a delay in hearing its own voice. The length of the delay depends on the delay inside the car and the Global System for Mobile Communications (GSM). The echo path length is a key parameter of AEC.
Figure 2: CVC-HFK application diagram
Adaptive filtering (more accurately, the NLMS algorithm) is one of the most common solutions for AEC. The NLMS algorithm achieves a good compromise between the amount of calculation and performance.
The other issue related to AEC is the fuzzy tone (DT) situation when the two are talking at the same time. If it is not detected, DT will cause the divergence of the adaptive algorithm.
The AEC software uses the NLMS algorithm to cancel the echo, which is executed by the C54x DSP assembler.
1. NLMS algorithm
The NLMS algorithm can update the coefficients of the adaptive finite impulse response (FIR) filter, which is used to predict the echo, and then we subtract the predicted value from the actual echo to give the residual echo.
2. Active channel detection
A key feature of the AEC algorithm is active channel detection. When the far-end operator is silent and the near-end operator speaks, because the near-end operator is no longer an echo, the filter does not need to be adapted. By calculating the signal energy and comparing this energy with the adaptive threshold, it can be achieved Detection of active channels.
3. Detection of fuzzy tone (DT)
In the case of DT, the near-end signal on the loudspeaker includes echo and near-end speech (ie, vague tone). The residual error used to update the filter coefficients includes near-end speech, and if the algorithm is still adapting, the algorithm may begin to diverge, and this must be avoided. DT detection uses an energy-based algorithm and a variable threshold to solve this problem.
4. Benchmark
The benchmark of the AEC software (expressed in 16-bit words) is:
Code size: 154 words;
Static RAM: 527 words;
Erase RAM: 2 words;
The maximum calculation cost is 4.7 MIPS. The computational consumption is the largest in the ST period, and drops to 2.4 MIPS in the DT period. The ST period occupies the main part of the call, while the DT period only appears in shorter individual cases.
CVC-HFK software
CVC-HFK (Clear Voice Capture-Hands Free Kit) integrates functions such as echo cancellation, noise suppression, and nonlinear processing, and is an optimized HFK solution. The CVC-HFK solution uses a comprehensive adaptive sub-band method to improve the performance of the main aspects while taking up very few resources. In the automotive environment, environmental noise is the main problem to be overcome by hands-free systems. Therefore, in addition to the echo cancellation function, Clarity CVC-HFK also provides an integrated single microphone solution (OMS) noise suppression algorithm. The OMS solution supports adaptive noise cancellation, which can attenuate the environmental noise in the microphone signal (incoming), extract the required speech, and transmit clean speech (outgoing) to the far-end user. Since CVC-HFK is fully adaptive, there is no need to adjust too much. Below, we will introduce the CVC-HFK solution and its main aspects of performance.
1. CVC-HFK AEC
The CVC-HFK echo canceller is a "stateless" AEC, which uses a variant of the standard frequency domain NLMS algorithm as its main adaptive filter. We will explain the benefits of using these methods below. First, the sub-band frequency domain method can uncorrelate or whiten the input signal in each band, which can achieve faster convergence compared to AEC in a considerable time domain. Second, stateless AEC enables continuous filter adaptation, which can improve robustness and overall fuzzy sound performance in noisy environments. As mentioned earlier, in the case of DT, the microphone signal contains both echo and near-end speech. The near-end voice is not associated with the echo signal, and if there is no process to avoid it, it will cause the divergence of the adaptive filter. Third, NLMS can achieve uniform convergence independent of input amplitude.
Because of this, CVC-HFK AEC can obtain a typical 40dB ERLE (echo return loss gain value), up to 50dB ERLE, and can achieve a fast convergence time of about 80ms, and can perform full-duplex operation in most environments . In addition, CVC-HFK AEC uses a tail length of 64ms for its adaptive filter, which allows greater flexibility in terms of internal capacity.
2. CVC-HFK NS (Noise Suppressor)
The CVC-HFK noise suppressor is a frequency domain algorithm that uses voice and noise characteristics to help extract voice from synthesized noise and voice signals. The two main modules of CVC-HFK NS are speech composition analysis and speech extraction.
The speech component analysis module uses the temporal and related attributes of speech and noise to construct a predictable model of speech composition. The speech extraction block can modify each frequency component according to the speech and noise model. In addition, the speech extraction block can also make full use of the principle of sound quality to minimize noise floor and perceived speech distortion.
CVC-HFK NS adopts this scheme to achieve 10-15dB SNR (signal-to-noise ratio) improvement in a noisy environment while maintaining good voice quality. In an extremely low noise environment where the SNR is already high enough, because the NS is turned off, no speech distortion will occur.
3. CVC-HFK NLP (non-linear processing)
As the system distortion increases, CVC-HFK NLP is minimal. The amount of distortion added by CVC-HFK NLP is much lower than standard NLP modules such as center clippers because it uses information from the input and error signals to determine additional attenuation. Since all CVC-HFK modules use frequency-domain algorithms, compared to solutions that use both time-domain and frequency-domain algorithms, they can significantly save memory and simplify computational complexity.
System integration design
When integrating a TI-HFK board with a cellular hands-free kit, several components and appropriate interfaces are required to achieve a good mobile call.
You must choose components that are compatible with both CVC-HFK application software and board hardware to achieve good performance. HFK can support various loudspeakers, loudspeakers and car audio systems. However, in order to reduce the changes to the application manual, we have selected dedicated industry standard components, which will greatly help your successful adjustment. Three connections are required from the TI-HFK board to the cellular kit to achieve integration: the loudspeaker inputs the TI HFK board; the outgoing, processed audio output; input, and the incoming signal received from the cellular kit.
Here are some suggestions for the design of loudspeaker and loudspeaker placement.
1. The position and orientation of the loudspeaker In order to achieve the best overall microphone performance, some key variables should be understood before the device is finally installed in the car. It is recommended to keep the distance between the microphone and the user's mouth in the car at 46cm (18 inches). The recommended distance ranges from 30 to 56cm (12-22 inches).
2. Avoid exposure of the loudspeaker to the airflow (windows and fans) as much as possible;
3. Appropriately consider the size and installation scheme of the loudspeaker so that the front of the loudspeaker can be aligned with the mouth of the user in the car. Based on the above considerations, with the help of Figure 3, you can choose the best loudspeaker position. First, follow the recommendations in Priority Zones 1, 2 and 3. Once a decision is made, you can fix the loudspeaker with a metal plate or velcro strap, and you can connect the cable back to the electronic device for termination. Afterwards, the concealment of the cable should be ensured to maintain its aesthetic appearance, and in addition, the cable should be kept fixed so as not to be pinched or knotted. Finally, avoid connecting parallel cables to antenna connectors and other noisy cables.
4. Speaker position It is recommended to install the speaker in an appropriate location to provide good voice performance without disturbing the pickup area of ​​the microphone. The pickup area is a +30 degree cone, which protrudes outward from the front of the microphone to the driver of the car.
Figure 3: Speaker placement
The speaker should be located at least 1 meter (3 feet) away from the microphone. The speaker should be far away from the pickup area of ​​the microphone to reduce the chance of echo feedback. Ideally, the speakers should be located behind the front of the microphone or at 90 degrees. The size of the speaker's sound distortion will have a direct negative impact on the echo.
Summary of this article
The use of hands-free systems continues to gain popularity, and users also expect that performance will continue to improve. Given the many options available for HFK implementation, it is obvious that integrating software algorithms and hardware signal processors is a thoughtful move, which will be very beneficial. The HFK development kit can cope with all the above problems and bring benefits to those who develop or sell the above products. The TI TMS320C5407 development kit with AEC and CVC-HFK provides the required flexibility and high performance, and can quickly and inexpensively bring HFK solutions to market.
The surface of charger gun is smooth and concise, handle grip is comfortable, and operation is convenient and safe. The charging gun meets IEC62196.2-2014 standard and has good interchangeability. The charging gun cable can be used for charging electric vehicles and can use Mode 2, Mode 3 to charge.
IEC Electric Vehicle Charging Plug
EV Plug,EV Charging Plug,EV Charging Outlet,Electric Car Socket
Suzhou WeBest Electronics Technology Co.Ltd , https://www.webestet.com