Variable Cutoff Frequency FIR Filters: A Survey

Many signal processing applications require digital filters with variable frequency characteristics, especially the filters with variable bandwidth. Due to their linear phase and inherent stability, variable bandwidth finite impulse response (FIR) filters are the popular choice in majority of the applications. Once a variable cutoff frequency (VCF) FIR lowpass filter is designed, variable bandwidth bandpass / highpass / bandstop filters and reconfigurable filter banks can be realized from the same. In this paper, we present a comprehensive review of the existing variable cutoff frequency FIR filter design techniques, including the developments in the recent two decades. We provide the basic concepts, design and architectural details for each of these techniques and the major developments / incremental works thereof. Qualitative as well as quantitative comparisons are provided to assist the reader in choosing the most suitable VCF filter design technique for a particular application.

💡 Research Summary

The paper presents a thorough survey of variable‑cutoff‑frequency (VCF) finite‑impulse‑response (FIR) filter design techniques, a topic that has gained increasing relevance as modern signal‑processing systems demand real‑time bandwidth adaptation. Beginning with a clear motivation, the authors explain why VCF FIR filters are preferred over IIR counterparts: they guarantee linear phase, unconditional stability, and straightforward re‑configurability, making them suitable for communication, radar, audio, and biomedical applications. The survey classifies existing methods into four principal families.

1. Coefficient Interpolation – Pre‑computed FIR coefficient sets for several fixed cut‑off frequencies are stored, and the desired response is obtained by linear, quadratic, or higher‑order interpolation between these sets. This approach is memory‑efficient and easy to implement in hardware, but interpolation errors can introduce ripple and phase distortion, especially for large frequency steps.

2. Frequency‑Transformation – A base low‑pass FIR filter is transformed digitally (band‑shift, scaling, or complex‑plane rotation) to achieve the required cut‑off. The technique preserves linear phase and stability while offering a wide tuning range, yet it adds extra multiplications and may require additional compensation filters to control pass‑band ripple.

3. Spectral Parameterization – Filter coefficients are expressed as explicit functions of the cut‑off frequency, typically via polynomial or rational models. Notable examples include Parametric FIR (PFIR) and Variable‑Coefficient FIR (VCF‑FIR). This family enables instantaneous parameter updates, making it ideal for fast‑changing environments, but higher‑order models increase design complexity and can suffer from numerical instability.

4. Multi‑Rate / Reconfigurable Filter Banks – By exploiting sample‑rate conversion and sub‑band decomposition, a small prototype filter is reused across multiple bands. This yields substantial savings in memory and power, which is attractive for ASIC or FPGA implementations targeting low‑power IoT devices. The downside is the need for careful alias‑ing and phase‑alignment handling at sub‑band boundaries.

For each category, the authors detail the theoretical foundation, typical architectural blocks, and representative publications spanning the last two decades. They provide side‑by‑side comparisons using quantitative metrics such as coefficient count, multiplier usage, pass‑band ripple, stop‑band attenuation, phase linearity error, and power consumption. Tables and plots illustrate how each method performs under different constraints (e.g., high‑speed vs. ultra‑low‑power).

A significant portion of the survey is devoted to hybrid techniques that combine strengths of multiple families—for instance, interpolated spectral‑parameter models or frequency‑transformed multi‑rate structures. These hybrids have been shown to mitigate individual weaknesses, delivering finer frequency resolution, lower computational load, and improved robustness. The paper also discusses design automation tools, optimization algorithms (genetic, particle‑swarm, and deep‑learning‑based parameter estimation), and hardware prototyping results on modern FPGA families and ASIC standard‑cell libraries.

Finally, the authors outline future research directions: ultra‑low‑power VCF FIR filters for edge AI sensors, simultaneous multi‑band adaptation for cognitive radio, and exploratory implementation in emerging quantum‑computing frameworks. In summary, the survey offers a comprehensive roadmap that equips designers with the knowledge to select the most appropriate VCF FIR technique for a given application, balancing performance, resource utilization, and implementation complexity.