Spectrum Prediction in the Fractional Fourier Domain with Adaptive Filtering

TL;DR

This paper introduces the SFFP framework that uses adaptive fractional Fourier transform and neural networks to improve spectrum prediction accuracy by better separating predictable patterns from noise.

Contribution

The paper proposes a novel spectrum prediction framework combining adaptive fractional Fourier transform, filtering, and complex neural networks, enhancing pattern separation and forecast accuracy.

Findings

SFFP outperforms existing spectrum forecasting methods.

Adaptive fractional Fourier transform improves pattern separability.

Neural network-based prediction achieves higher accuracy.

Abstract

Accurate spectrum prediction is crucial for dynamic spectrum access (DSA) and resource allocation. However, due to the unique characteristics of spectrum data, existing methods based on the time or frequency domain often struggle to separate predictable patterns from noise. To address this, we propose the Spectral Fractional Filtering and Prediction (SFFP) framework. SFFP first employs an adaptive fractional Fourier transform (FrFT) module to transform spectrum data into a suitable fractional Fourier domain, enhancing the separability of predictable trends from noise. Subsequently, an adaptive Filter module selectively suppresses noise while preserving critical predictive features within this domain. Finally, a prediction module, leveraging a complex-valued neural network, learns and forecasts these filtered trend components. Experiments on real-world spectrum data show that the SFFP outperforms leading spectrum and general forecasting methods.