This work employs modular quantum machine learning (QML) to estimate the wireless channels in simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) aided communication systems. Although RISs, composed of low-energy phase-shifting elements, can enable controlled signal reflections to cover communication devices obstructed by blockages, the devices located behind the reflection-only surfaces cannot be covered as these structures now become blockages themselves. STAR-RISs solve this issue by allowing the transmission signals to be conveyed to the devices located behind the STAR-RIS structures. However, acquiring accurate channel information of devices in the reflection and transmission regions of a STAR-RIS is not a trivial task. To address this issue, this paper proposes a novel modular QML scheme that employs different quantum-based learning modules to (i) eliminate the noise from the coarse channel information, and (ii) estimate the channels of the devices in the reflection and transmission regions.
This work employs modular quantum machine learning (QML) to estimate the wireless channels in simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) aided communication systems. Although RISs, composed of low-energy phase-shifting elements, can enable controlled signal reflections to cover communication devices obstructed by blockages, the devices located behind the reflection-only surfaces cannot be covered as these structures now become blockages themselves. STAR-RISs solve this issue by allowing the transmission signals to be conveyed to the devices located behind the STAR-RIS structures. However, acquiring accurate channel information of devices in the reflection and transmission regions of a STAR-RIS is not a trivial task. To address this issue, this paper proposes a novel modular QML scheme that employs different quantum-based learning modules to (i) eliminate the noise from the coarse channel information, and (ii) estimate the channels of the devices in the reflection and transmission regions.
A modular quantum machine learning (QML) based framework for channel estimation and tracking in STAR-RIS aided wireless systems, is proposed. While STAR-RISs enable extensive 360-degree manipulation of the propagation channels compared to reflection-only RISs, it is not a trivial task to acquire and track channel information in STAR-RIS aided systems, as this needs to be done for the devices located in both the reflection zone and the transmission zone of the surface. Powered by quantum-based operations, which are characterized by quantum parallelism, QML holds the potential to handle high-dimensional channel estimation and tracking in STAR-RIS systems. Towards this end, a novel modular QML framework that employs different quantum-based learning modules is proposed. Its purpose, threefold in design, strives to (i) eliminate the noise inherited in the coarse channel information, (ii) estimate the channels of devices in the reflection and transmission regions, and (iii) update the estimated channels to accommodate the time-varying movements of these devices. The numerical results demonstrate that the proposed QML-based approach outperforms the one based on classical recurrent neural networks.
Sixth-generation (6G) wireless networks are expected to deliver transformative capabilities including ultra-reliable low-latency communications and massive connectivity. Reconfigurable intelligent surfaces (RISs) have emerged as a key enabling technology by dynamically manipulating wireless propagation through programmable reflecting elements. However, accurate channel estimation in RIS-assisted networks faces significant challenges from cascaded channel structures, passive elements, and high-dimensional parameter spaces. Quantum machine learning (QML) offers a fundamentally different computational paradigm that may address these challenges through quantum properties. This article provides a tutorial overview of QML for channel estimation in RIS-assisted 6G networks. We present RIS-assisted system architectures and channel estimation challenges, introduce quantum computing fundamentals, and survey QML architectures. As a use case, we consider a hybrid quantum-classical framework, e.g., convolutional neural network (CNN)-quantum long short-term memory (QLSTM), for channel estimation in RIS-assisted non-orthogonal multiple access (NOMA) systems. This case study shows consistent accuracy improvements compared to quantum and classical benchmarks. Finally, we discuss implementation challenges and future research directions for practical deployment of quantum-enhanced wireless systems.
This study proposes the utilization of quantum machine learning (QML) to maximize the energy efficiency of reconfigurable intelligent surface (RIS) assisted communication with rate-splitting multiple access (RSMA). The next-generation wireless communications are expected to yield significantly higher energy efficiency compared to that of the previous generations. In a multiuser system, energy efficiency can be defined as a benefit-to-cost ratio between the achievable sum-rate and the energy consumption, where enhancements in the former come at the expense of increases in the latter. Recently, the integration between RSMA and RISs has been advocated as a powerful mean to control this tradeoff. Indeed, RSMA can enhance the rate region while RISs can lead to reduced energy consumption thanks to the use of low-energy phase shifters. However, optimizing a RIS-aided RSMA communication system is faced with a computational burden given that the RIS enlarges the volume of the required channel information, which expands the information that needs to be processed by the optimization module, even when the optimization is based on conventional learning techniques. The proposed QML optimization framework, which orchestrates non-linear quantum unitary operations to compose the learning models, enjoys information processing gains thanks to state vector operations in multi-dimensional Hilbert space. It is composed of two trainable quantum-based learning models employed in an alternating manner: the first establishes the transmission precoding, and the second designs the RIS phase shifting. Numerical results show that the proposed QML delivers comparable performance to that of conventional optimization but with reduced complexity.
Recently, there has been a growing interest in employing reconfigurable intelligent surfaces (RISs) to improve the spectrum and energy efficiency of wireless networks. In RIS-assisted wireless networks, channel estimation and optimization is a difficult task, particularly for nearly passive RIS devices with low-complexity hardware design. We have proposed a hybrid classical-quantum computing framework that allows for ultra-fast optimization adapting to multipath wireless environments. The onsite optimization of RIS configuration can be performed almost instantaneously using only feedback (received power) at wireless endpoints. The performance of the proposed work is demonstrated in representative wireless propagation scenarios.
A reconfigurable intelligent surface (RIS) is a promising technology for beyond-fifth-generation (B5G) and sixth-generation (6G) wireless communications, but its passive reflection and two-hop double-fading structure make cascaded channel estimation challenging. Conventional convolutional neural network (CNN) estimators require many trainable parameters, while a single shallow parameterized quantum circuit (PQC) may have limited feature representation. Deep quantum circuits can also suffer from noise and barren-plateau effects on noisy intermediate-scale quantum (NISQ) devices. To address these issues, this paper proposes a multi-filter quantum convolutional neural network (MF-QCNN) for cascaded channel estimation in RIS-assisted multi-user uplink systems. The proposed model uses multiple independent shallow PQC filters in parallel, concatenates their measured features, and estimates the cascaded channel through a compact classical dense head, with the total trainable-parameter count scaling as 182F+696 for F parallel filters. Simulation results, compared with a single-filter quantum convolutional neural network (QCNN), CNN, and multilayer perceptron (MLP) baselines, show that at a signal-to-noise ratio (SNR) of 20 dB, the 3-filter MF-QCNN reduces the normalized mean squared error (NMSE) by approximately 22.9, 8.1, and 4.6 dB relative to the single-filter QCNN, CNN, and MLP baselines, respectively, while using only about 19.3% of the CNN trainable parameters. Under zero-forcing (ZF) precoding, it achieves the highest achievable sum rate among the learning-based estimators; at SNR = 30 dB, it improves the achievable sum rate by approximately 17.4% and 12.8% over the CNN and MLP baselines, respectively. These simulation results suggest that the parallel shallow-PQC design can serve as a compact quantum-aided estimator for RIS channel estimation and may provide a useful basis for future studies on AI-native transceiver design in B5G/6G networks.
Simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) enables 360-degree full-space coverage by controlling both transmitted and reflected signals through each element. However, the coupled phase-shift model, where the transmission and reflection coefficients are physically interdependent, transforms the phase optimization into an extremely complex non-convex combinatorial problem that classical algorithms struggle to solve efficiently. In this paper, we propose a hybrid quantum-classical framework that employs the Quantum Approximate Optimization Algorithm (QAOA) to optimize the coupled phase-shift configuration of STAR-RIS. Each discrete phase configuration is mapped to a qubit-based Ising Hamiltonian, enabling QAOA to explore the exponentially large solution space through quantum superposition and entanglement. To address the scalability challenge of simulating large antenna arrays (e.g., 8 × 8 = 64 elements), we introduce a variable grouping (tiling) strategy that partitions the RIS into manageable sub-arrays of 2 × 2 elements, each requiring only 8 qubits for 2-bit phase quantization. Our comprehensive simulations demonstrate that the proposed QAOA-based optimizer achieves an approximation ratio exceeding 0.94 with 10 QAOA layers, maintains over 86% fidelity under 2% gate error rates in noisy intermediate-scale quantum (NISQ) environments, and exhibits robust performance under imperfect channel state information with less than 5% degradation at 30% CSI error variance. The energy efficiency analysis confirms that the QAOA-optimized STAR-RIS achieves approximately 1.7× higher spectral efficiency per unit energy compared to conventional reflect-only RIS. Furthermore, the proposed tiling framework scales linearly with array size, enabling deployment on arrays up to 16 × 16 elements while maintaining constant per-group quantum resource requirements.