Implementation method and system for DSP multi-channel howling reduction processing and chip

The invention discloses an implementation method and system for DSP specified multi-channel howling reduction processing and a chip. The method comprises the following steps: a DSP audio processor generates at least one signal source with balanced test to a specified output sound box to be played in an external environment for feedback acquisition processing; an FFT detection control module in communication connection with the DSP audio processor obtains an equalization signal source of a feedback sound field in the environment from at least one input microphone, and a calibration sound sourceobtained after feedback of the environment and the whole system is obtained for spectral analysis; and equalization processing is carried out on the equalization quantity value in the spectral analysis result, and the frequency response curve of the environment where the corresponding microphone and the corresponding sound box are located is controlled through the DSP audio processor, so that thefrequency point with high