Designing Quantum Gradient Descent Algorithm for MIMO NOMA Rate Maximization With STAR-RIS

• Published in: IEEE wireless communications letters Vol. 14; no. 4; pp. 959 - 963
• Main Authors: Paul, Anal, Singh, Keshav, Li, Chih-Peng, Mumtaz, Shahid
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.04.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Array signal processing; Beamforming; Communication networks; Downlink; gradient descent algorithms; Linear programming; NOMA; Nonorthogonal multiple access; Optimization; Parallel processing; Quantum computing; quantum gradient descent; Reconfigurable intelligent surfaces; Reflection; search space-time complexity; Signal to noise ratio; Vectors; Wireless communications
• ISSN: 2162-2337; 2162-2345
• DOI: 10.1109/LWC.2025.3528382

We introduce, for the first time in wireless communication networks, a quantum gradient descent (QGD) algorithm to maximize sum data rates in non-orthogonal multiple access (NOMA)-based simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-assisted multiple-input and multiple-output systems. The QGD algorithm utilizes the principles of quantum parallelism and superposition to efficiently solve the high-dimensional optimization challenges inherent in configuring transmit and receive beamformers and STAR-RIS elements. Extensive simulations demonstrate that the QGD algorithm significantly outperforms classical optimization methods, achieving up to 49.50% and 44.88% higher data rates compared to classical gradient descent algorithms for configurations with 256 STAR-RIS elements. Furthermore, the NOMA model shows substantial improvements in sum data rate performance, with gains of 179.65% and 145.61% over space division multiple access schemes under similar frameworks.