Apparatus and method for implementation of fixed decimal sampling frequency conversion

Abstract

The invention discloses a method and a device which can realize decimal sample-rate conversion, and the method is used in the data sampling conversion field. The method includes the steps that an integral-multiple interpolation filter is used to conduct integral-multiple interpolation and filtering on an inputted signal and an output integral-multiple signal to a decimal sample-rate converter; the decimal sample-rate converter conducts decimal sample-rate conversion on a received decimal audio video signal, and outputs a decimal conversion audio video signal to an integral-multiple extraction filter; the integral-multiple extraction filter obtains the converted sample rate after conducting integral-multiple extraction and filtering on a received integral-multiple conversion audio video signal, and the decimal sample-rate conversion of the ratio of the output signal to the input signal is realized. With the technical proposal of the invention, when the interpolation multiple I or the extraction integer D of the decimal sample-rate conversion is bigger, the noise rate of the output signal can be improved, thereby the noise rate performance of the output single can be met.

Description

technical field [0001] The invention relates to data sampling conversion, in particular to a method and device capable of realizing fixed fractional sampling rate conversion. Background technique [0002] When a multimedia terminal receives or sends audio and video digital signals, it often needs to perform fractional sampling rate conversion processing on the audio and video digital signals; in another multi-mode digital mobile communication device, since the symbol rate of each mode is not an integer multiple, it is also A fractional sample rate conversion is required. The implementation process of the currently commonly used fractional sampling rate conversion method is to set the ratio of the output sampling rate to the input sampling rate as I / D, where I and D are natural numbers that are relatively prime. Please refer to accompanying drawing 1, general implementation method comprises the following steps: [0003] 101. Interpolation of I times is carried out to the in...

AI Technical Summary

Problems solved by technology

[0013] However, this method is effective when the values ​​of I and D are small, but when the value of I or D is relatively large, and the SNR (Signal to Noise Ratio, noise rate) of the signal is required to be high, a large amount of filtering needs to be stored If the device coefficient is low, the system hardware will be difficult to realize;

[0014]The second method is to use interpolation approximation methods, such as linear interpolation and polynomial interpolation; US Patent Specification US5907295 discloses a method using FIR low-pass fil

Images