Frequency-Shifting Digital Filter for Narrow-Band Signal Suppression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Communication receivers face challenges in achieving good selectivity at low implementation cost for suppressing narrow-band signals, as existing digital filters are either costly or inefficient in filtering out unwanted signals.

Innovation Solution

A digital filter design incorporating frequency shifting mechanisms, signal processors, and decimators/interpolators to efficiently notch unwanted frequency ranges while maintaining low implementation costs, utilizing a combination of frequency shifting, filtering, and signal processing to shift and filter input signals effectively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional digital filters are used to suppress narrow-band signals, then selectivity is improved, but implementation cost increases

Engineering Contradiction:
ImproveselectivityVSAvoidimplementation cost
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The filter is divided into multiple parallel filter banks, each handling a specific frequency range. This segmentation allows each sub-filter to operate independently with reduced complexity, achieving high overall selectivity without requiring a single complex filter structure

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the filtering problem from the time domain to the frequency domain by using FFT-based filter banks. This dimensional transformation enables efficient frequency-selective filtering through simple multiplication in the frequency domain, reducing computational complexity compared to traditional time-domain filtering

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If complex filtering algorithms are used to achieve good selectivity, then signal suppression performance is improved, but computational effort increases

Engineering Contradiction:
Improvesignal suppression performanceVSAvoidcomputational effort
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces complex time-domain filtering operations with frequency-domain multiplication operations using FFT. This substitution transforms computationally intensive convolution operations into simple point-wise multiplications in the frequency domain, dramatically reducing computational effort while maintaining filtering performance

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent uses adjustable frequency shift parameters to dynamically reposition the filter banks. By changing the frequency domain parameters rather than redesigning the entire filter structure, the system achieves adaptive signal suppression with minimal computational overhead

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP1976121A1Digital filter
Publication Date: 2008.10.01 SONY DEUT GMBH
  • EP1976121A1 patent drawingFigure 1A
  • EP1976121A1 patent drawingFigure 1B
  • EP1976121A1 patent drawingFigure 1C

AI summary

Digital filter (104), comprising: a first frequency shifting mechanism (106) configured to shift an input spectrum (H(f)) of an input signal (100) by a first frequency distance (F1) to obtain a shifted frequency signal (122) with a shifted spectrum; a filter (108), configured to filter said shifted frequency signal with a predetermined transfer function to obtain a filtered signal (130); and a signal processor (110), said signal processor including: an adder (202), and at least a second frequency shifting mechanism (222), said signal processor (110) being configured to generate an output signal (140), which is a sum of said filtered signal (130) and said shifted frequency signal (122) and said output signal (140) has an output spectrum, which is shifted by a second frequency distance (DF) with respect to said shifted spectrum, wherein said second frequency distance (DF) is independent of said first frequency distance (F1).