Research On Bluetooth Audio Signal Processing Algorithm

Bluetooth technology has been developed for 22 years since its birth.Due to its continuous absorption of new technologies,proposes new functions and expands new application environments,Bluetooth technology has a strong vitality.The Advanced Audio Distribution Profile(A2DP)in the Bluetooth standard defines a subband codec(SBC).As a codec that must be supported in Bluetooth audio device,it is of great significance to improve and optimize the implementation algorithm of this codec to improve the quality of Bluetooth audio.This paper focuses on the improvement and optimization of the SBC codec algorithm.On the basis of being familiar with the Bluetooth protocol stack and the Bluetooth audio application development framework,this paper research and analyze the principle of filter bank and modulator algorithm in the SBC codec,and optimize the algorithm for the problems of poor codec quality and low data compression rate in the current algorithm.Based on the completion of the algorithm optimization,the hardware design of each module of the SBC codec is completed,and finally the RTL hardware circuit of the SBC codec is obtained.The main research results and innovations of this paper are as follows:1.Improve the cosine modulation formula.Aiming at the problems of poor subband division performance and serious distortion of the original filtering system,the constraint relationship between the analysis filter and the synthesis filter is analyzed in the design of the filtering system by combining the Quadrature Mirror Filter Bank design method,and it is applied to the Cosine Modulation Filter Bank design method,obtain an improved cosine modulation formula,which eliminates phase distortion and most of aliasing distortion.Compared with the original cosine modulation formula,the filter bank obtained by improving the cosine modulation formula improves the passband bandwidth of the low-frequency and highfrequency subbands by 68.14%,while reducing the amplitude distortion by 39.10%,which can effectively improve the audio quality of the low frequency and high frequency parts of the audio signal.2.Optimize the design of prototype low-pass filter.In the design of the prototype low-pass filter,based on the Kaiser window function method,this paper introduces the filter transition band and stopband attenuation indicators as constraint functions,and obtains the real coefficient linear phase FIR low-pass filter by iterative calculation.The designed filter has a high stopband attenuation and a reasonable transition band change trend.When designing a 4 subband filter bank,after 10 iterations,a prototype filter with an order of 41 and a minimum stopband attenuation of-65.53 d B is obtained.When designing an 8 subband filter bank,the prototype filter with order of 79 and minimum stopband attenuation of-116.62 d B is obtained after 21 iterations.The resulting real coefficient cosine-modulated filter bank has lower noise and better subband analysis and synthesis performance than the filter bank in the original method.3.Improve adaptive differential pulse code modulation.For the Adaptive Pulse Code Modulation(APCM)technology used in the SBC codec algorithm,it can only be limited to adjust the coding quantization and coding data volume larger problems,this paper combines the Differential Pulse Coded Modulation(DPCM)technology to improve the original method of directly coding and modulating the input signal to code and modulate the difference between the input signal and its predicted value.At the same time,a nonlinear compression algorithm is added to the original modulation method,so that the encoder can evenly encode the large and small signals in the audio signal.Compared with the original modulator,the Adaptive Differential Pulse Coded Modulator(ADPCM)after comprehensive optimization has a data compression rate of up to 6: 1,and the encoding and decoding quality is higher.4.Design and implement the SBC codec hardware circuit.This paper provides the hardware implementation of the improved SBC codec and gives the RTL circuit design.According to the characteristics of the subband codec,a parallel method is used to design the hardware circuit to improve the operation speed of the codec and reduce the algorithm delay.At the same time,a CRC check module is added to the codec to effectively ensure the correctness of the codec.