Summary

This two-part blog shows how polynomial evaluation can be mapped onto FIR filter engines, explaining why the structural similarity makes filters a natural substrate for polynomials and where the approach breaks down. Part 1 focuses on the basic mapping, implementation caveats (dynamic range, quantization, accumulator use) and applications such as modeling transducer and amplifier nonlinearity.

Key Takeaways

• Map polynomial evaluation onto FIR/filter-engine structures using Horner-like decomposition and block-FIR techniques.
• Analyze trade-offs in dynamic range, accumulator width, and coefficient quantization when using dedicated filter hardware.
• Design fixed-point scaling and overflow-handling strategies suitable for real-time FIR-based polynomial calculation.
• Apply the technique to model and compensate transducer and amplifier nonlinearity in audio, radar, and communications chains.

Who Should Read This

Embedded DSP engineers and algorithm designers with some experience in filter implementations who want to exploit FIR hardware to compute polynomials efficiently for audio, radar, or communications applications.

Still RelevantIntermediate

Topics