Voice uploading (talking on the phone): Voice is received by the microphone as a low-frequency analog signal, converted into a digital signal by an analog-to-digital converter, and processed by digital signal processing such as data compression, cyclic double check code, channel coding, interleaving, encryption, formatting, multiplexing and modulation through a "baseband chip".
Next, it is converted into a high-frequency digital signal (electromagnetic wave) by an "intermediate frequency chip (IF)", namely a high-frequency analog-to-digital converter (DAC); Finally, the "RF chip" forms electromagnetic waves with different time, frequency and waveform, which are emitted by the antenna.
Voice download (phone call): the antenna receives electromagnetic waves with different time, frequency and waveform, and after being processed by "RF chip", it gets high-frequency digital signals (electromagnetic waves), which are then converted into digital signals by "IF chip", that is, high-frequency digital-to-analog converter (ADC).
Next, demodulation (demodulation), demultiplexing, reformatting, decryption, deinterleaving,
Digital signal processing, such as channel decoding, cyclic double check code (CRC), data decoding, etc. Finally, it is converted into a low-frequency analog signal (sound) by a low-frequency digital-to-analog converter (DAC) and played by a microphone.
Extended data:
Signal generation
DTMF encoder is based on two second-order digital sine wave oscillators, one for generating line frequency and the other for generating column frequency. By loading the corresponding coefficients and initial conditions into DSP, eight audio signals can be generated with only two oscillators.
The frequency range of a typical DTMF signal is 700~ 1700Hz. Choosing 8000Hz as the sampling frequency can satisfy the Nyquist condition. DTMF dual audio signals are generated by two second-order digital sinusoidal oscillators, one for generating line audio signals and the other for generating column audio signals.
Generation process
CCITT stipulates that the maximum number of keys per second is 10, that is, the shortest time slot of each key is 100MS, in which the actual duration of audio is at least 45MS, not more than 55MS, and the rest time slots are silent, so when a key generates a dual audio signal, two consecutive signals are separated for a period of time; The decoder uses this time to identify the dual audio signal and convert it into corresponding digital information.
But also gap information. Therefore, this process includes audio tasks and silent tasks. The former generates double audio samples, while the latter generates silent samples. At the end of each task, the timer and the next task should be reset. Among them, the silent task also adds a task: take out the numbers from the digital buffer and unpack them. Unpacking is to map numbers to corresponding line and column audio features,
The loading pointer points to the correct position corresponding to the oscillator characteristic table. These two tasks take turns. According to CCITT (International Telegraph and Telephone Advisory Committee), there must be a proper length of silence between numbers, so the encoder has two tasks, one is to generate an audio signal with two-tone samples, and the other is to generate a mute sample. After each mission,
Before starting the next task (audio signal task or mute task), the timer variable that determines its duration must be reset. After the mute task is completed, DSP calls up the next number from the digital cache and makes a judgment. Line frequency and column frequency signals corresponding to these signals, and their initialization parameters are determined according to different frequencies.
identify
DTMF signal contains two sets of audio signals, and the task of the decoder is to transform it from time domain to frequency domain through mathematical transformation, and then get the corresponding digital information. Because the chip processes digital signals, it is necessary to digitize the input signals and then process them with DSP chips.
In frequency detection, the fundamental wave and second harmonic of DTMF signal are detected. DTMF signal only has high energy on the fundamental wave, while speech signal has strong second harmonic superimposed on the fundamental wave. The function of detecting the second harmonic is to distinguish DTMF signals from language and music signals.
References:
Baidu encyclopedia -DTMF