Dataset for Power Control in Cell-Free Massive MIMO

Description

This repository contains a dataset that represents the simulation of 12 distinct cell-free massive MIMO scenarios. Each scenario includes 20,000 setups, derived from a customized version of the MATLAB code package available here. The simulations take into account three power control optimization schemes: max-min Spectral Efficiency (SE) fairness, sum SE maximization, and Fractional Power Control (FPC).

The results for the simulation for each scenario are split across 20 MATLAB storage files ('power_control_part#.mat'), with each file encompassing 1,000 setups. Denoting the number of access points (APs) as L, the numer of user equipments (UEs ) as K, and the number of setups as M. The variables within these files are defined as follows:

• 'gainOverNoise': This is a [L x K x M] matrix that contains the large-scale fading coefficients.
• 'pk_UL_nopt_LPMMSE_X': This is a [K x M] matrix that includes the power coefficients with; (1) X = 'maxmin' for the max-min SE fairness, (2) X = 'sumSE' for the sum SE maximization, and (3) X = 'FPC' for the FPC power scheme.
• 'SE_UL_nopt_LPMMSE_X': This is a [K x M] matrix that includes the SE per UE with; (1) X = 'full' for all the UEs transmitting with maximum power, (2) X = 'maxmin' for the max-min SE fairness, (3) X = 'sumSE' for the sum SE maximization, and (4) X = 'FPC' for the FPC power scheme.
• 'bk_dist_X': This is a [K x M] matrix that represents the average channel gain of the desired signal with; (1) X = 'frac' for the FPC, and (2) X = 'full' for the rest of power control schemes.
• 'ck_dist_X': This is a [K x K x M] matrix the represents the average channel gains for the respective interfering signals; (1) X = 'frac' for the FPC, and (2) X = 'full' for the rest of power control schemes.
• 'sigma2_dist_XX': This is a [K x M] matrix the represents the effective noise variance; (1) X = 'frac' for the FPC, and (2) X = 'full' for the rest of power control schemes.

Additonally, the file 'setup.mat' contains information about the simulated environment.

Please refer to the conference paper above for more detailed information about the simulations.

The simulation code is available here on GitHub for training and testing the DNN models.

Referencing

If you in any way use this dataset for research that results in publications, please cite our original article listed above.

G. García-Barrios, M. Fuentes, and D. Martín-Sacristán, “A Flexible Low-Complexity DNN Solution for Power Control in Cell-Free Massive MIMO,” in 2024 IEEE 35th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 2024, pp. 1–6.

Acknowledgements

This work is supported by the grant from the Spanish ministry of economic affairs and digital transformation and of the European Union – NextGenerationEU [UNICO-5G I+D/AROMA3D-Earth] (TSI-063000-2021-69).