Fractional FIR Sample Rate Conversion for Non-Integer Resampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sample rate conversion technologies are inflexible and require hard-coded resamplers, limiting the ability to convert signals to non-integer multiples of the initial sampling rate, and often necessitate additional filters and complex processing.
Innovation Solution
A system utilizing a fractional finite impulse response (FIR) circuit with a phase accumulator and look-up table (LUT) to dynamically adjust coefficients, enabling conversion to any desired sampling rate, including non-integer multiples, with parallel processing to enhance efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If hard-coded resamplers are used for sample rate conversion, then the conversion process is simple, but the system lacks flexibility and cannot convert to non-integer multiples of the initial sampling rate
Solution Approach 1:
The patent implements a dynamic resampler where the filter coefficients are not fixed but are dynamically selected based on the desired sampling rate conversion ratio. The system uses a coefficient selector that chooses appropriate coefficients from a set of pre-calculated values based on the conversion ratio R, allowing the same hardware structure to adapt to different sampling rate conversions including non-integer multiples.
Solution Approach 2:
The patent changes the parameter of filter coefficients from fixed to variable. By providing a set of pre-calculated coefficients corresponding to different conversion ratios and selecting the appropriate coefficients based on the desired R value, the system achieves flexibility in sampling rate conversion without requiring complex real-time coefficient calculation.
2Adaptability or versatility
If additional filters are added to achieve flexible sample rate conversion, then the conversion capability is improved, but the device complexity increases
Solution Approach 1:
The patent creates a universal resampler structure that can handle multiple sampling rate conversion scenarios (upsampling, downsampling, and non-integer conversions) using a single unified architecture. The same FIR filter structure with dynamically selected coefficients serves multiple conversion purposes, eliminating the need for separate dedicated filters for each conversion type.
Solution Approach 2:
Instead of adding multiple fixed filters for different conversion ratios, the patent changes the coefficients of a single filter structure dynamically. By storing multiple sets of coefficients and selecting the appropriate set based on the conversion ratio R, the system achieves versatile conversion capability without increasing the fundamental filter structure complexity.
3Measurement precision
If complex processing is used to achieve precise sample rate conversion, then the conversion accuracy is improved, but the processing speed decreases
Solution Approach 1:
The patent performs complex coefficient calculations in advance and stores the results in a lookup table. During actual signal processing, the system only needs to select pre-calculated coefficients based on the conversion ratio rather than computing them in real-time. This preliminary preparation of coefficient data enables both high conversion precision and fast processing speed.
Solution Approach 2:
The system dynamically selects from pre-computed coefficients based on the conversion ratio, allowing precise interpolation and filtering operations to be performed efficiently. The dynamic coefficient selection enables the system to adapt to different conversion requirements while maintaining both precision and processing speed.
Data Source
AI summary
An example system is configured to convert a first digital signal having a first sampling rate into a second digital signal having a second sampling rate that is different from the first sampling rate. The system includes an input circuit to receive the first digital signal at the first sampling rate; a fractional finite impulse response (FIR) circuit configured to shift the first digital signal by a value corresponding coefficients of the FIR circuit, with the value being based on an integer value or a non-integer value; memory to store the coefficients for the FIR circuit; and processing circuitry to receive information corresponding to the second sampling rate, to obtain the value based on the information, to obtain the coefficients from the memory based on the value, and to provide the coefficients to the FIR circuit.