Beamforming Design for Active STAR-RIS-Aided CF-mMIMO Networks Under Non-Linear Hardware Impairments and Electromagnetic Interference

Owing to the inherent abilities of active simultaneously transmitting and reflecting reconfigurable intelligent surfaces (aSTAR-RIS) to simultaneously reflect, refract, and amplify the incident signals, it is being used to enhance the user experience in a wireless network. In this work, we consider aSTAR-RIS to enhance the energy efficiency (EE) of cell-free massive multiple-input multiple-output (CF-mMIMO) network in the presence of hardware impairments (HWI) and electromagnetic interference (EMI). We aim to jointly optimize the respective transmit and reflect beamformers at the access points (APs) and aSTAR-RIS and formulate an EE maximization problem. The maximization problem is solved by leveraging fractional programming tools to decouple the transmit and reflect beamformers and, subsequently, solve it using an alternating optimization (AO) approach. To reduce the computational complexity of AO, we derive closed-form updates for the HWI-aware transmit beamformers at the APs and resort to stochastic projected gradient descent method with closed-form step size updates to optimize the reflect beamformers at aSTAR-RIS. We also provide an effective AP selection scheme to reduce the fronthaul signaling and further enhance the EE of the network. Simulation results corroborate the effectiveness of the proposed approach in maximizing the EE of the considered network.