A New Class of Structured Beamforming for Content-Centric Fog Radio Access Networks

TL;DR

This paper introduces a novel regularized zero forcing beamforming method for content-centric fog radio access networks, reducing computational complexity and optimizing energy efficiency while managing interference and multiple constraints.

Contribution

It proposes a new class of beamforming that simplifies decision variables and develops low-complexity algorithms for energy-efficient optimization in F-RANs.

Findings

Effective mitigation of inter-content interference.

Reduced computational complexity for beamformer design.

Enhanced energy efficiency under multiple constraints.

Abstract

A multi-user fog radio access network (F-RAN) is designed for supporting content-centric services. The requested contents are partitioned into sub-contents, which are then 'beam- formed' by the remote radio heads (RRHs) for transmission to the users. Since a large number of beamformers must be designed, this poses a computational challenge. We tackle this challenge by proposing a new class of regularized zero forcing beamforming (RZFB) for directly mitigating the inter-content interferences, while the 'intra-content interference' is mitigated by successive interference cancellation at the user end. Thus each beamformer is decided by a single real variable (for proper Gaus- sian signaling) or by a pair of complex variables (for improper Gaussian signaling). Hence the total number of decision variables is substantially reduced to facilitate tractable computation. To address the problem of energy efficiency optimization subject to multiple constraints, such as individual user-rate requirement and the fronthauling constraint of the links between the RRHs and the centralized baseband signal processing unit, as well as the total transmit power budget, we develop low-complexity path- following algorithms. Finally, we actualize their performance by simulations.