Group Sparse Precoding for Cloud-RAN with Multiple User Antennas.

Area of Science:

• Wireless Communications
• Signal Processing
• Network Engineering

Background:

• Cloud Radio Access Network (C-RAN) architecture is crucial for next-generation cellular networks due to massive data traffic.
• Group sparse beamforming in C-RAN enhances spectral efficiency by activating subsets of remote antenna ports (RAPs).
• Existing schemes often overlook the complexity of multiple-antenna users, necessitating joint optimization of transmit and receive beamformers.

Purpose of the Study:

• To formulate an optimal joint RAP selection and precoding design for C-RAN with multiple-antenna users.
• To investigate the trade-off between sum rate and the number of active RAPs under fixed transmit power constraints.
• To develop an efficient algorithm for this complex optimization problem.

Main Methods:

• Formulation of the group sparse precoding problem using an L0-norm penalty, addressed via a reweighted L1 heuristic.
• Leveraging block diagonalization precoding to transform the problem into a convex optimization.
• Development of an efficient algorithm based on the Lagrangian dual of the convex problem.

Main Results:

• The proposed algorithm effectively handles joint optimization of RAP selection and precoding for multiple-antenna users.
• Achieved sum rates closely approximate those from exhaustive search methods.
• Demonstrated an efficient approach to a previously complex problem in C-RAN beamforming.

Conclusions:

• The developed algorithm provides an efficient and effective solution for group sparse precoding in C-RAN with multiple-antenna users.
• This work advances C-RAN design by enabling high spectral efficiency in complex multi-antenna environments.
• The findings support the practical implementation of advanced beamforming techniques in future cellular networks.