Multi-Stage FFT Engine Data Normalization for Buffer Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern wireless communication systems face challenges in reducing the size of FFT engine output buffers due to the large dynamic range of received OFDM signals, which necessitates significant memory usage and increases computational resources.
Innovation Solution
Implementing a series of stages in the FFT engine with butterflies and data normalization devices that scale output signals, allowing for bit-width reduction and smaller buffer sizes by normalizing the signals between stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the FFT engine processes received OFDM signals without data normalization, then the dynamic range of the output signal is maintained, but the buffer size and computational resources increase significantly
Solution Approach 1:
The FFT processing is divided into multiple stages, with data normalization applied between stages. This segmentation allows the system to manage the dynamic range in increments rather than handling the full range in a single operation, reducing the buffer size required at each stage while maintaining overall dynamic range performance.
Solution Approach 2:
The patent changes the parameter of signal amplitude by applying data normalization (scaling) between FFT stages. This parameter change reduces the dynamic range of intermediate results, allowing for smaller buffer sizes and reduced computational resources while the final stage restores the original dynamic range to maintain performance.
2Quantity of substance
If the FFT engine uses multiple stages with data normalization, then the buffer size and computational resources are reduced, but the device complexity increases
Solution Approach 1:
By dividing the FFT engine into multiple stages with normalization operations between them, the patent reduces the buffer size required at each stage. The segmentation creates manageable intermediate results that require less memory, offsetting the increased structural complexity with reduced resource requirements.
3Reliability
If the FFT engine processes signals with large dynamic range, then the signal quality is maintained, but the computational operations increase
Solution Approach 1:
The patent applies parameter changes by normalizing the signal amplitude between FFT stages. This reduces the dynamic range of intermediate computational operations, decreasing the number of computational operations required while maintaining signal quality through proper scaling at each stage and restoration at the final stage.
Data Source
AI summary
In embodiments, a fast Fourier transform (FFT) engine includes a series of stages, each stage containing a butterfly and a data normalization device configured to scale output of the stage's butterfly. The scaling factors are adjusted, for example, periodically or on as-needed basis, so that the dynamic range of the butterflies and the buffers is increased for a given bit-width, or the bit-width of these devices is decreased for the same dynamic range. Additionally, bit-width of other buffer(s) is decreased because of the scaling of the data.