Mode-Switching Discrete-Phase Shifters

Updated 27 October 2025

Mode-switching discrete-phase shifters are hardware components that toggle between finite phase states to enable efficient beamforming and adaptive surface configurations in mmWave and 6G systems.
They integrate binary and multi-bit phase control with switching networks to reduce hardware complexity, power consumption, and enable scalability in hybrid MIMO and IRS architectures.
Advanced optimization and channel estimation techniques, including MINLP and machine learning methods, enhance performance while supporting near-ideal transmission and adaptive reflection in next-gen networks.

Mode-switching discrete-phase shifters are hardware elements capable of toggling between quantized phase states or operational modes to control electromagnetic signals in reconfigurable surfaces and beamforming arrays. Unlike fully analog, continuously adjustable phase shifters, mode-switching discrete-phase shifters feature either binary selection (on/off, transmit/reflect) or operate over a finite set of phase values (typically 1–4 bits resolution). These devices are increasingly employed in millimeter-wave (mmWave) and extremely-large-scale (XL-array) systems to address power, cost, and scalability constraints, while enabling near-ideal beamforming, adaptive transmission–reflection switching, and low-complexity hybrid MIMO architectures.

1. Architectures and Hardware Principles

Mode-switching discrete-phase shifters are implemented within mmWave hybrid beamforming systems, intelligent reflecting/transmitting surfaces (IRS/STAR-RIS), photonic processors, and time-modulated phased arrays. Typical architectures combine discrete phase shifters with switching networks:

Architecture	Phase Control	Switch Role
Subarray Partitioning	Discrete/Fixed	Connects subsets, enables mode
Fully-Connected Switch	Adaptive/Discrete	Any antenna to any RF chain
Binary Switch Network	1-bit/On-off	Low-complexity selection
Mode-Selective Shifter	Spatial mode control	Differentiates modes (e.g. TE0/TE1)

Switches are used to select which antennas or subarrays participate in combining, while discrete-phase shifters restrict each signal path to a set of quantized phase values (e.g., {0, π/2, π, 3π/2} for 2-bit). In STAR-RIS and IRS, elements can switch between transmission and reflection operational modes, each with dedicated phase and amplitude settings (Alishahi et al., 20 Oct 2025).

2. Optimization Frameworks and Joint Design

Optimization involving mode-switching discrete-phase shifters is intrinsically mixed-integer and nonlinear due to discrete and binary constraints. Contemporary solutions employ block coordinate descent with tailored subproblem solvers:

Mixed-integer Nonlinear Programming (MINLP) handles variables such as user power allocation, active beamforming matrices, binary switches (αₙ^{t/αₙ^r} ∈ {0,1}), and discrete phase vectors (θₙ^X ∈ {0,2π/Q,…,2π(Q–1)/Q}) (Alishahi et al., 20 Oct 2025).
Block Coordinate Descent (BCD):
- Power allocation via Difference-of-Concave (DC) programming, exploiting objective decomposition into concave components for iterative refinement.
- STAR-RIS phase optimization via closed-form updates, followed by quantization (projecting onto nearest phase value).
- Binary amplitude optimization via combinatorial methods (branch-and-bound, MOSEK).
Combinatorial Optimization: Selects mode (transmit/reflect) and corresponding phase from a finite alphabet for each element.

These approaches efficiently manage the combinatorial explosion induced by discrete switches and phase quantization.

3. Impact on Performance, Energy Efficiency, and Scalability

The combination of mode switching and phase discretization introduces distinct trade-offs in system metrics:

Spectral Efficiency: Moderate phase quantization (e.g., 3–4 bits) can approach the performance of continuous phase shifters, as shown in sum rate maximization for STAR-RIS (Alishahi et al., 20 Oct 2025).
Energy Efficiency: Discrete-phase shifters and switch networks consume significantly less power than their analog counterparts, especially critical in mobile terminals and XL-arrays (Méndez-Rial et al., 2015, Zhang et al., 23 Sep 2024).
Hardware Complexity: The adoption of binary switching and quantized phase reduces component count and cost, with architectures supporting large antenna arrays via grouped connections (Payami et al., 2017, Qi et al., 2022).
Grating Lobes: Discrete phase control inherently introduces additional grating lobes (type-I and type-II), attenuating peak beam power and potentially degrading communication rates, though moderate quantization (≥3 bits) effectively suppresses these artifacts (Zhang et al., 23 Sep 2024).

4. Channel Estimation and Beamforming Algorithms

Discrete-phase hardware mandates robust estimation and beamforming solutions:

Open-loop Compressive Channel Estimation: Compressive sensing-based estimators reconstruct sparse multipath channels independent of analog implementation (switch or shifter-based), using sparse recovery methods such as OMP in hybrid architectures (Méndez-Rial et al., 2015).
Training Matrix Optimization: In IRS-aided channel estimation, discrete unitary matrices (Hadamard or DFT) ensure minimal MSE with low-resolution phase shifters, with 1-bit sufficing for Hadamard (Sun et al., 2021).
Machine Learning-based Reflections: Cross-entropy algorithms adapt IRS phase settings within their discrete sets, alternating with ZF-based transmit precoding, to minimize BS power while satisfying SINR constraints (Yan et al., 2022).
Heuristic and Greedy Selection: Incremental antenna selection and phase shifter activation maintain spectral efficiency with reduced hardware.

5. Applications in Active and Passive Surfaces

Mode-switching discrete-phase shifters are foundational in the following systems:

STAR-RIS: Simultaneous transmit-and-reflect reconfigurable intelligent surfaces leverage binary amplitude and discrete phase for cost and performance efficiency, with mixed-integer joint optimization for multi-user sum-rate design (Alishahi et al., 20 Oct 2025).
IRS and RIS: Active (beamforming at BS) and passive (IRS phase setting) beamforming co-optimization is effective in 6G scenarios, with mode switching enabling both reflective and transmissive propagation control.
Photonic Mode Selectivity: Silicon photonics chips equipped with mode-selective thermo-optic phase shifters independently control multiple spatial modes (e.g., TE0/TE1), supporting mode-division multiplexing and quantum optical processors (Safaee et al., 2023).
Time-Modulated Phased Arrays: Sequential switching of multi-bit phase states yields single-sideband modulation and agile beam steering with reduced hardware complexity (Gao et al., 2020).

6. Practical Design Guidelines

Empirical and theoretical results inform several practical design rules:

Quantization Resolution: A phase shifter resolution of 3–4 bits achieves near-ideal beamforming and channel estimation, with diminishing returns for higher bit counts (Sun et al., 2021, Zhang et al., 23 Sep 2024).
Mode Allocation: Minimum allocation constraints per mode (e.g., N/3 elements per transmission/reflection) assure coverage and system robustness in STAR-RIS deployments.
Power Consumption Models: Switch-consumption (∼5 mW) is much lower than phase shifters (∼30 mW for 4–5 bits), favoring switch-based design for low-power terminals (Méndez-Rial et al., 2015, Payami et al., 2017).
Complexity Management: Closed-form alternation and heuristic update rules (thresholding, SVD-based updates) prevent design bottlenecks in large-scale architectures (Yu et al., 2017, Qi et al., 2022).
Grating Lobe Mitigation: Moderate bit quantization and digital beamforming substantially reduce pattern degradation and interference due to grating lobes.

7. Outlook and Implications for 6G IoT and Beyond

Mode-switching discrete-phase shifters underpin scalable, energy-conserving architectures for next-generation wireless and optical networks:

STAR-RIS and IRS-assisted MIMO designs support sum-rate maximization with low-cost, energy-aware hardware, readily scaling to dense IoT deployments and XL-array systems (Alishahi et al., 20 Oct 2025, Zhang et al., 23 Sep 2024).
Photonic chips leveraging mode-selectivity enhance quantum processing and reconfigurable optics (Safaee et al., 2023).
Adaptation in near-field and far-field regimes, via quantization-aware strategies (DTPQ/EIPQ), ensures robust beamforming performance under diverse propagation scenarios (Sang et al., 2023).
Algorithmic frameworks combining block coordinate descent, combinatorial optimization, and machine learning efficiently navigate the discrete design space, enabling practical deployment of complex, reconfigurable surfaces.

The persistent direction is for architectures that couple binary mode-switching and multi-bit phase quantization, reducing cost and complexity yet preserving high performance—meeting the demands of emerging 6G systems and energy-aware networks.