2025, Vol. 6, Issue 1, Part A

Statistical signal processing plays a pivotal role in the design and optimization of modern wireless communication systems, covering fundamental tasks such as estimation, detection, filtering, and inference under uncertainty ^{[1, 2]}. As wireless communication evolves from 4G to 5G and toward 6G, the wireless channel environment has become increasingly complex, with challenges arising from dynamic fading, large-scale multiple-input multiple-output (MIMO) systems, millimetre-wave (mmWave) and terahertz (THz) bands, reconfigurable intelligent surfaces (RIS), and the growing demands of ultra-dense connectivity and low latency ^[3-4]. Traditional signal processing frameworks, although robust in stable, low-dimensional settings, struggle to accommodate non-stationarity, high-dimensionality, and the deep integration of machine learning algorithms. The central problem explored in this paper “Statistical Signal Processing in Wireless Communications: Challenges and Solutions” is how to extend statistical algorithms to remain scalable, robust, and efficient in next-generation wireless systems.

This paper aims to:

(i) Identify and analyse key challenges in statistical signal processing for emerging wireless technologies (such as channel estimation under high mobility, interference mitigation in massive MIMO, blind detection in multi-user settings, and signal processing in RIS- or THz-assisted communications),

(ii) Evaluate state-of-the-art solutions such as Bayesian Monte Carlo methods, sequential Monte Carlo techniques, and machine learning-based unfolded frameworks ^[5-6], and

(iii) Present a hypothesis that statistical methods incorporating non-stationarity, high-dimensional stochastic models, and machine-learning elements will outperform classical approaches in key wireless metrics.