Adaptive Sample Rate Conversion Using NCO-Driven Filter Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sample rate conversion systems face complexity and limitations in accommodating non-integer factors, requiring large lookup tables and being non-adaptive, especially when dealing with fractional baud rates and timing skew in digital demodulation processes.
Innovation Solution
The implementation of a sample rate converter using a Numerically Controlled Oscillator (NCO) to generate filter coefficients based on input and output frequencies, allowing for a wide range of non-integer factors, and incorporating parallel filter taps to accommodate fractional baud rates with fixed clock hardware, while using sign-magnitude format for multiplication and truncation/rounding schemes to manage precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large lookup tables are used for sample rate conversion, then conversion accuracy is improved, but device complexity increases
Solution Approach 1:
The lookup table is segmented into multiple smaller tables, each storing coefficients for specific frequency ranges or conversion ratios. This segmentation reduces the memory access complexity and allows for more efficient retrieval, maintaining conversion accuracy while reducing overall device complexity
Solution Approach 2:
The system dynamically selects which lookup table to use based on the current input frequency and desired output frequency. This dynamic selection mechanism allows the system to maintain high conversion accuracy across a wide range of frequencies without requiring all possible conversion coefficients to be stored simultaneously, thereby reducing device complexity
2Device complexity
If fixed clock hardware is used, then hardware requirements are simplified, but adaptability to fractional baud rates decreases
Solution Approach 1:
The system changes the parameters of the filter coefficients stored in the lookup tables to accommodate different baud rates including fractional rates. By pre-calculating and storing coefficients for various fractional baud rate scenarios, the fixed clock hardware can adapt to different data rates without requiring variable clock frequencies, thus maintaining simple hardware requirements while improving adaptability
Solution Approach 2:
The lookup table structure is designed to be universal, containing filter coefficients that can be applied across multiple baud rate scenarios including integer and fractional rates. This multi-functional approach allows a single fixed clock hardware implementation to serve multiple data rate requirements, enhancing adaptability without increasing hardware complexity
Data Source
AI summary
Generally described herein are methods and systems for sample rate conversion of non-integer and integer factors. In one or more embodiments an apparatus can include a sample rate converter that can include an input configured to receive an input signal with a first frequency and an output configured to provide an output signal with a second frequency different from the first frequency. The sample rate converter can include a filter coefficient lookup table and a numerically controlled oscillator configured to provide filter coefficients from the filter coefficient lookup table at a rate that is a function of the first frequency and the second frequency. The sample rate converter can include a multiplier configured to produce an output that is the product of a filter coefficient of the filter coefficients from the numerically controlled oscillator and a sample of an input signal and an accumulator configured to sum an output of the multiplier and provide a result of the summation when the accumulator receives an indicator to dump the result.