Dataset Title: Beamforming Matrix for Satellite Communications

Dataset Description: Beamforming coefficients to generate a specific beam pointing towards a specific point, with specific beamwidth, specific EIRP, and specific Side-Lobe Level (SLL).

Efficient antenna design plays a pivotal role in communication satellite missions, particularly due to substantial free-space losses and stringent restrictions on allowable losses. To tackle these challenges, our team has innovatively crafted an open waveguide antenna, described in detail in [1]. This antenna design boasts significantly reduced losses compared to alternative solutions like patch antennas or dielectric-based antennas. It consists of three primary components: the open-ended waveguide, the slot polarizer for circular polarization, and the rectangular-to-circular transition for connection to the distribution network.

Our approach centers on array thinning, involving the activation and deactivation of elements to generate the desired beam. This method empowers us to create controlled and variable beamwidths, effectively distribute power among elements in multi-beam scenarios, and orient the radiation pattern as needed through progressive phase shifts.

[1] J. A. Vásquez-Peralvo et al., “Flexible Beamforming for Direct Radiating Arrays in Satellite Communications,” in IEEE Access, vol. 11, pp. 79684-79696, 2023, doi: 10.1109/ACCESS.2023.3300039.
[2] F. Ortiz, J.A. Vasquez-Peralvo, J. Querol, E. Lagunas, L. Garces, J.L.  Gonzalez-Rios, V. Monzon, S. Chatzinotas, “Harnessing Supervised Learning for Adaptive Beamforming in Multibeam Satellite Systems”, submitted to IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN), Stockholm, Sweden, May 2023
[3] F. Ortiz, J.A. Vasquez-Peralvo, J. Querol, E. Lagunas, J.L. Gonzalez-Rios, M. Oliveira, L.M. Garces-Socarras, V. Monzon-Baeza, S. Chatzinotas, “Supervised Learning Based Real-Time Adaptive Beamforming On-Board Multibeam Satellites”, submitted to European Conference on Antennas and Propagation (EuCAP), Glasgow, Scottland, March 2024.
[4] A. Vasquez-Peralvo, J. Querol, E. Lagunas, F. Ortiz, M. Oliveira, L.M. Garces-Socarras, J.L. Gonzalez-Rios, S. Chatzinotas, “Genetic Algorithm-based Beamforming in Subarray Architectures for GEO Satellites”, submitted to European Conference on Antennas and Propagation (EuCAP), Glasgow, Scottland, March 2024.

Database Objective:

The primary objective of this database is to establish a correlation between beamforming matrix weights and key system parameters, including beamwidth, Side Lobe Level (SLL), and Effective Isotropic Radiated Power (EIRP). We generated this database through the application of a Genetic Algorithm (GA). Users provide input parameters such as beamwidth, EIRP, SLL, and beam scanning angles to the algorithm. The GA then identifies the beamforming matrix that most closely aligns with the specified input parameters.

Database Contents:

Within this database, you will find seven distinct datasets, each providing valuable information:

1. Input Algorithm: This dataset comprises the input data fed into the GA. It takes the form of an Nx4 matrix, with ‘N’ representing the number of samples generated. The four columns correspond to the desired parameters for each iteration, following this order: EIRP, beamwidth, azimuth, and elevation.

2. Output Algorithm: This dataset contains real values obtained when assessing the beamforming matrices derived from the GA. It consists of an Nx6 matrix, with six parameters obtained in each iteration, as follows: EIRP, beamwidth, azimuth, elevation, SLL, and the number of active elements within the beamforming matrix.

In addition to these datasets, we provide the actual beamforming matrix for each sample, identified by a sample number within the “outputMatrix” name. In total, this comprehensive database comprises a substantial 174,203 samples. It has been curated with the intention of facilitating the application of Machine Learning models for further analysis and optimization.

DOWNLOAD DATA