Deep Unfolded Fractional Optimization for Maximizing Robust Throughput in 6G Networks

TL;DR

Addresses robust downlink beamforming under imperfect CSI by maximizing the robust weighted sum rate $R_{ ext{max}}^{\gamma}$ using an uncertainty-injected deep unfolded fractional programming (UI-DUFP) framework. The method unfolds FP iterations into $M$ trainable layers with $N$ PGD steps, while injecting channel uncertainty across $B$ realizations to optimize a quantile-based objective. Results show UI-DUFP achieves higher robust WSR and lower inference time than WMMSE, FP, and DL baselines, with 8PGD offering additional gains, demonstrating scalable, uncertainty-aware optimization for real-time 6G networks.

Abstract

The sixth-generation (6G) of wireless communication networks aims to leverage artificial intelligence tools for efficient and robust network optimization. This is especially the case since traditional optimization methods often face high computational complexity, motivating the use of deep learning (DL)-based optimization frameworks. In this context, this paper considers a multi-antenna base station (BS) serving multiple users simultaneously through transmit beamforming in downlink mode. To account for robustness, this work proposes an uncertainty-injected deep unfolded fractional programming (UI-DUFP) framework for weighted sum rate (WSR) maximization under imperfect channel conditions. The proposed method unfolds fractional programming (FP) iterations into trainable neural network layers refined by projected gradient descent (PGD) steps, while robustness is introduced by injecting sampled channel uncertainties during training and optimizing a quantile-based objective. Simulation results show that the proposed UI-DUFP achieves higher WSR and improved robustness compared to classical weighted minimum mean square error, FP, and DL baselines, while maintaining low inference time and good scalability. These findings highlight the potential of deep unfolding combined with uncertainty-aware training as a powerful approach for robust optimization in 6G networks.

Figures (5)

• Figure 1: Network setup with a multi-antenna BS serving 4 users, illustrating the imperfect channel knowledge.
• Figure 2: Flow diagram of the proposed UI-DUFP framework illustrating the input parameters (channel matrix and initial beamforming matrix), the optimized beamforming output, and the quantile-based robust WSR.
• Figure 3: Robust WSR vs. number of unfolded layers $M$ (UI-DUFP) and iteration count (WMMSE and FP) with $\sigma^2_h = 0.05$.
• Figure 4: Robust WSR as a function of $\sigma^2_h$, comparing UI-DUFP (4PGD and 8PGD) with WMMSE, FP, and DL-UI baselines.
• Figure 5: Inference time versus number of antennas and users ($L=K$) for all evaluated schemes.