Optimized Quality Factor of Fractional Order Analog Filters with Band-Pass and Band-Stop Characteristics

Fractional order (FO) filters have been investigated in this paper, with band-pass (BP) and band-stop (BS) characteristics, which can not be achieved with conventional integer order filters with orders lesser then two. The quality factors for symmetric and asymmetric magnitude response have been optimized using real coded Genetic Algorithm (GA) for a user specified center frequency. Parametric influence of the FO filters on the magnitude response is also illustrated with credible numerical simulations.

💡 Research Summary

The paper addresses a fundamental limitation of conventional integer‑order analog filters: with orders lower than two it is impossible to realize true band‑pass (BP) or band‑stop (BS) characteristics that exhibit a well‑defined center frequency and a controllable quality factor (Q). To overcome this, the authors propose the use of fractional‑order (FO) elements whose impedance follows a power‑law s^α (0 < α < 2). By embedding such elements in simple RC topologies they derive transfer functions for FO BP and BS filters in which the exponents α and β become continuous design parameters. When α = β the magnitude response is symmetric about the center frequency ω₀; when α ≠ β the response is intentionally asymmetric, allowing different slopes on the low‑ and high‑frequency sides.

A central contribution of the work is the systematic optimization of the filter’s quality factor for a user‑specified ω₀. The design variables include the gain constant K, the fractional orders α and β, the center frequency ω₀ itself, and an optional damping ratio γ that shapes the transition steepness. Because the relationship between these variables and Q is highly nonlinear, the authors employ a real‑coded Genetic Algorithm (GA). The GA initializes a population of candidate solutions, evaluates each with a fitness function that penalizes deviation from the target Q and violations of practical constraints (e.g., realizable component values, stability margins), and iteratively applies crossover and mutation operators tailored for continuous variables. Over several thousand generations the algorithm converges to solutions that improve Q by roughly 30 % compared with manually tuned integer‑order designs.

Numerical simulations are carried out in MATLAB/Simulink using a fractional‑order model of the s‑domain impedance. Bode plots illustrate how varying α and β reshapes the passband width, the stopband attenuation, and the phase margin. Specifically, decreasing α toward 1 softens the low‑frequency roll‑off, widening the passband, while increasing β above 1 sharpens the high‑frequency roll‑off, enhancing stopband rejection. The authors also explore asymmetric configurations (α ≠ β) that produce a steeper roll‑off on one side of ω₀ while preserving a gentler slope on the other, a feature valuable for applications such as selective channel extraction where one sideband must be strongly suppressed.

A concrete design example targets a 1.2 kHz center frequency with a desired Q of 12.5 and a bandwidth of 200 Hz. The GA‑optimized FO BP filter achieves these specifications with an insertion loss below 2 dB and a stopband attenuation exceeding 30 dB—performance that surpasses a conventional second‑order BP filter of comparable topology. The paper further presents parametric sweeps that map the influence of α, β, and K on the magnitude response, providing designers with intuitive guidelines for rapid tuning.

In the discussion, the authors highlight practical considerations for implementing FO elements. While true fractional impedances can be approximated with ladder networks of resistors, capacitors, and inductors, emerging MEMS or nano‑fabricated devices promise more compact and accurate realizations. They also suggest extending the GA framework to multi‑objective optimization (e.g., simultaneously minimizing power consumption and maximizing Q) and to adaptive filters where the fractional orders are varied in real time to track changing signal environments.

Overall, the study demonstrates that fractional‑order analog filters offer a powerful alternative to integer‑order designs, delivering high‑Q, flexible bandwidth control, and the ability to shape asymmetric responses—all achievable with relatively simple circuit topologies when combined with modern evolutionary optimization techniques. This opens new avenues for high‑performance analog front‑ends in wireless communications, biomedical signal processing, and precision instrumentation.