Source-linked AI summary
Intelligent Surfaces for 6G Wireless Networks: A Survey of Optimization and Performance Analysis Techniques
Rawan Alghamdi, Reem Alhadrami, Dalia Alhothali, Heba Almorad, Alice Faisal, Sara Helal, Rahaf Shalabi, Rawan Asfour, Noofa Hammad, Asmaa Shams, Nasir Saeed, Hayssam Dahrouj, Tareq Y. Al-Naffouri, Mohamed-Slim Alouini
TL;DR
LIS research requires coordinated optimization and performance analysis to assess its role in future wireless networks. This survey combines physical principles, optimization frameworks, performance analyses, positioning applications, and open challenges.
Problem
LIS systems require analysis of optimization objectives, communication performance, hardware impairments, positioning, and emerging deployment concerns.
Method
The paper surveys LIS operating principles alongside optimization methods for energy efficiency, power, sum-rate, secrecy-rate, and coverage, plus performance-analysis frameworks.
Results
The survey covers capacity, hardware-impairment effects, uplink and downlink rates, outage probability, positioning, and open challenges for LIS-aided wireless networks.
Takeaways & Limitations
LIS are presented as a physical-layer technology that can enhance wireless quality of service while reducing power consumption relative to traditional networks.
Takeaways & Limitations
Some surveyed optimization methods provide estimated or locally optimal solutions rather than guaranteed global optima.
Abstract
from arXiv · showhide
This paper surveys the optimization frameworks and performance analysis methods for large intelligent surfaces (LIS), which have been emerging as strong candidates to support the sixth-generation wireless physical platforms (6G). Due to their ability to adjust the behavior of interacting electromagnetic (EM) waves through intelligent manipulations of the reflections phase shifts, LIS have shown promising merits at improving the spectral efficiency of wireless networks. In this context, researchers have been recently exploring LIS technology in depth as a means to achieve programmable, virtualized, and distributed wireless network infrastructures. From a system level perspective, LIS have also been proven to be a low-cost, green, sustainable, and energy-efficient solution for 6G systems. This paper provides a unique blend that surveys the principles of operation of LIS, together with their optimization and performance analysis frameworks. The paper first introduces the LIS technology and its physical working principle. Then, it presents various optimization frameworks that aim to optimize specific objectives, namely, maximizing energy efficiency, sum-rate, secrecy-rate, and coverage. The paper afterwards discusses various relevant performance analysis works including capacity analysis, the impact of hardware impairments on capacity, uplink/downlink data rate analysis, and outage probability. The paper further presents the impact of adopting the LIS technology for positioning applications. Finally, we identify numerous exciting open challenges for LIS-aided 6G wireless networks, including resource allocation problems, hybrid radio frequency/visible light communication (RF-VLC) systems, health considerations, and localization.
I. INTRODUCTION
LIS are presented as programmable surfaces that control electromagnetic propagation, offering potential gains in wireless connectivity, efficiency, and 6G applications. The survey combines physical principles, optimization frameworks, performance analysis, and open research directions.
- Motivation: LIS can reduce power consumption because their nearly passive metasurfaces do not require analog/digital converters or power amplifiers.The paper connects this low-noise, energy-efficient operation with reduced environmental impact.
- Applications: LIS support emerging applications across indoor and outdoor smart environments, including IoT, D2D communication, sensing, and high-frequency communications.The paper also discusses integration with NOMA and full-duplex communication.
- Motivation: LIS use software-controlled surfaces to shape electromagnetic propagation and create more controllable radio environments.They are described as complementary to technologies such as mMIMO, backscatter, millimeter-wave communication, and network densification.
- Survey scope: The survey is positioned as the first work combining mathematical optimization and performance analysis of LIS systems.It covers optimization objectives including energy efficiency, power, sum-rate, secrecy-rate, and coverage, alongside capacity, rate, outage, and positioning analyses.
- Working principle: The physical mechanism uses reconfigurable metamaterials whose induced phase shifts can be adjusted to control electromagnetic reflection.The paper explains that meta-atom structure determines whether incoming waves are absorbed or reflected.
- Working principle: The received destination power is directly proportional to N^2, demonstrating a power gain with the number of independently controlled LIS phases.The same relationship is described as inversely proportional to the square of source-destination distance.
III. OPTIMIZATION USE CASES IN IRS-BASED SYSTEMS
IRS optimization addresses the rising data-rate and energy demands of beyond-5G wireless systems. The surveyed systems use controllable phase shifts to improve objectives such as energy efficiency, while avoiding additional power amplifiers.
- Optimization motivation: IRS optimization targets data rate, power consumption, energy efficiency, and secrecy rate in beyond-5G wireless systems.IRS can adjust each reflecting element’s phase shift to constructively combine reflected signals.
A. Energy Efficiency
Energy-efficiency optimization in LIS systems targets the trade-off between throughput and power consumption by jointly configuring transmit resources and IRS phase shifts. The surveyed approaches address practical constraints including discrete phase shifts, QoS requirements, and non-convex coupling.
- Energy-efficiency motivation: Energy efficiency balances wireless throughput against power consumption, motivating optimized IRS strategies.IRS phase shifts can constructively combine reflected signals without additional power amplifiers.
- Optimization formulation: The EE formulation includes individual QoS constraints, a BS power budget, and discrete IRS phase-shift constraints.The resulting problem is non-convex and is addressed using alternating optimization and gradient-based approaches.
- Practical phase control: Discrete phase shifts provide a more practical alternative to ideal continuous phase-shift models under hardware limitations.One approach alternately tunes individual phase shifts while fixing the remaining elements.
- SWIPT: IRS-assisted SWIPT optimization jointly addresses information decoding and energy harvesting by maximizing weighted harvested power subject to SINR and power constraints.Alternating optimization and semidefinite relaxation are used, and reported results show improved rate-energy performance.
- Received-power optimization: Passive-IRS received-power maximization can be reformulated through semidefinite relaxation and solved as a convex semidefinite program.A distributed low-complexity algorithm can avoid both AP–IRS feedback and SDP solutions when closed-form phase updates exist.
C. Sum-rate Maximization
Sum-rate optimization studies how active beamforming, passive IRS reflection, phase-dependent amplitudes, and learning-based control can improve achievable rates or spectral efficiency. Because these formulations are generally non-convex, surveyed methods include alternating optimization, SDR, manifold optimization, and fixed-point iteration.
- Learning-based rate optimization: Deep learning is used to select IRS beamforming from a codebook when exhaustive search is too complex for practical implementation.The objective is to find an optimal achievable rate under quantized beamforming constraints.
- Spectral-efficiency maximization: Manifold optimization and fixed-point iteration are reported to achieve higher spectral efficiency with lower computational complexity.The underlying phase-shift optimization problem is non-convex.
- Relaxation and optimality: SDR reformulates the phase-shift problem as a QCQP relaxation, but discarding the rank-one constraint yields an estimated rather than guaranteed optimal solution.Fixed-point iteration and manifold optimization can instead seek locally optimal solutions.
- Hardware-aware modeling: Practical reflection models account for phase-dependent amplitudes rather than assuming one reflection amplitude for every phase shift.The resulting non-convex problem is solved with alternating optimization to obtain a sub-optimal solution.
- Weighted sum-rate: Weighted sum-rate maximization jointly optimizes BS transmit beamforming and IRS reflection coefficients under ideal, continuous, or discrete phase-shifter models.The formulation includes user priorities and a maximum transmit-power constraint.
D. Secrecy-rate Maximization
Secrecy-rate optimization uses IRS phase shifts and transmit covariance or beamforming design to improve legitimate communication relative to interception. The surveyed formulations alternate between transmit-side and IRS-side optimization under unit-modulus and power constraints.
- Security objective: IRS secrecy optimization seeks to increase the legitimate receiver’s rate while decreasing the eavesdropper’s rate.The surveyed system includes Alice, Bob, Eve, and an IRS with controllable reflecting elements.
- Joint formulation: The secrecy-rate problem jointly optimizes the transmit covariance matrix and IRS phase-shift matrix under transmit-power and unit-modulus constraints.The achievable secrecy rate is the objective to maximize.
- Solution strategy: Alternating optimization fixes the IRS phase matrix while optimizing the transmit matrix, then reverses the roles.The unit-modulus constraints make the problem non-convex.
- Phase optimization: Fractional programming is used for the IRS phase-update subproblem, with extensions to multi-antenna eavesdroppers.The multi-antenna extension considers Eve with M ≥ 1 antennas.
E. Coverage Optimization
Coverage-oriented optimization includes max-min SINR design for IRS-assisted multi-user systems. The surveyed analyses use phase optimization, optimal linear processing, closed-form solutions in special channels, and deterministic approximations for higher-rank settings.
- Max-min SINR: The max-min SINR objective designs the IRS phase matrix to improve the weakest user’s downlink SINR.The formulation is built from the per-user downlink SINR and precoding model.
- Optimal linear processing: Optimal linear processing can solve the max-min SINR problem by allocating powers while designing the IRS element phases.The cited formulation assumes infinite-resolution phase shifters and perfect channel knowledge at the BS.
- Channel-rank cases: For a rank-one LoS BS–IRS channel, a closed-form minimum-SINR solution is available under optimal linear processing.As the user count increases, serving multiple users becomes more challenging because of SINR convergence.
- Channel-rank cases: For high-rank LoS channels, random matrix theory provides deterministic approximations for the optimal linear-processing parameters.The survey places this work within broader IRS optimization frameworks covering multiple objectives and assumptions.
IV. PERFORMANCE ANALYSIS OF LIS SYSTEMS
Performance analysis of LIS systems is important because their large dimensions make closed-form mathematical descriptions challenging. The reviewed analyses therefore address data rates, outage probability, and spectral efficiency.
- Large LIS dimensions make closed-form performance expressions challenging to obtain.
- Reviewed performance analyses cover uplink and downlink data rates, outage probability, and spectral efficiency.
A. Capacity Analysis of LIS Systems
Capacity analyses examine normalized capacity and spatial degrees of freedom for infinitely sized LIS deployments under fixed transmit power per unit volume. They show that user spacing determines capacity-maximizing spatial multiplexing densities.
- Capacity analysis considers an infinitely sized LIS with fixed transmit power per unit volume as wavelength approaches zero.
- Spatial degrees of freedom are measured through normalized capacity and the high-SNR slope to align LIS analysis with Shannon capacity.
- For one-dimensional deployment, normalized capacity is maximized when 2/λ UEs are multiplexed per meter.
- For two-dimensional deployment, normalized capacity is maximized when π/λ^2 UEs are multiplexed per square meter.
- A hexagonal lattice minimizes LIS surface area under the stated sampling condition.
2) Hardware Impairments Analysis:
Hardware impairments can reverse the usual capacity benefit of enlarging an LIS surface, while asymptotic analyses characterize rate behavior and channel hardening in single- and multi-unit systems.
- Hardware Impairments Analysis: Hardware impairment variance is modeled as a function of distance from the LIS center, with α = 0 representing no hardware impairment.
- Hardware Impairments Analysis: Increasing LIS surface area severely degrades capacity under hardware impairments, so the surface can be split into K smaller units.
- Uplink Rate for single-LIS Systems: Asymptotic LIS data-rate evaluation can avoid extensive simulations by using the asymptotic mean and variance of γ_k.
- Uplink Rate for single-LIS Systems: For increasing N, the LIS interference-plus-noise term and data-rate mean and variance converge to constants, while rate variance approaches zero.
2) Uplink Rate for Multi-LIS Systems:
Multi-LIS uplink analysis derives asymptotic spectral-efficiency bounds, studies pilot contamination, and optimizes training and served-device parameters. Practical phase-shift limitations are also linked to attainable-rate degradation and required phase-shift resolution.
- Uplink Rate for Multi-LIS Systems: Multi-LIS analysis derives an upper bound on asymptotic system spectral efficiency and investigates pilot contamination.
- Uplink Rate for Multi-LIS Systems: The multi-LIS framework optimizes pilot-training length and the number of served devices per LIS while modeling CSI acquisition and inter-LIS interference.
- Uplink Rate for Multi-LIS Systems: Pilot contamination bounds multi-LIS spectral efficiency through inter- and intra-interference caused by line-of-sight paths.
- Uplink Rate for Multi-LIS Systems: For LIS systems, increasing training length does not increase SINR, so the SINR-maximizing choice is the minimum t = K.
- Rate Impact on Phase Shifts: Limited phase shifts require quantized values si,jΔθ and can reduce reliability relative to continuous phase shifts.
- Rate Impact on Phase Shifts: When the LIS is sufficiently large, SNR is proportional to the square of the number of LIS elements, while phase-shift bit requirements depend on the system setting.
V. RELIABILITY ANALYSIS OF LIS
Reliability analysis examines outage probability and error performance in LIS systems. The surveyed results use asymptotic rate distributions and phase-aware transmission models to characterize reliability under different operating conditions.
- Rate Distribution and Outage Probability: Outage probability analysis is needed because individual user rates are not identically distributed, making the sum-rate distribution non-trivial.The surveyed approach instead models each rate through a random variable and estimates the sum-rate distribution for large N and K.
- Rate Distribution and Outage Probability: For large N and K, the sum-rate distribution can be approximated as Gaussian, enabling a closed-form outage-probability expression.The expression uses the sum-rate threshold R_D and the Q-function.
- Probability of Error for Intelligent and Blind Transmission: Intelligent transmission with phase adjustment achieves low error probability even at low SNR values.The cited study compares intelligent transmission, blind transmission, and LIS-AP operation.
- Probability of Error for Intelligent and Blind Transmission: Doubling N improves error performance by 6 dB under intelligent transmission.Blind transmission obtains an N × SNR gain over point-to-point transmission, while LIS-AP intelligent transmission improves by 1 dB.
C. Phase Shift Error Effect on Transmission
This section studies how phase uncertainty affects LIS transmission, including phase estimation and quantization errors, reflection geometry, and surface size. The surveyed analyses show that phase errors alter channel statistics but do not eliminate important LIS scaling benefits.
- Phase Shift Error Effect on Transmission: Practical phase shifts are chosen to cancel the combined source-to-reflector and reflector-to-destination channel phases, maximizing SNR.The model considers phase estimation and quantization errors for a finite number of reflectors.
- Phase Shift Error Effect on Transmission: With phase errors, the LIS channel is equivalent to a point-to-point Nakagami-fading channel whose parameters depend on the first two circular moments.The phase-uncertainty model treats reflector phase errors through their characteristic function.
- Phase Shift Error Effect on Transmission: With phase errors, average SNR increases with N^2 and diversity order increases with N.Numerical error-rate results indicate robust LIS performance even with a limited number of reflectors.
- Reflection Probability of LIS Systems: Object reflection probability depends on LIS length and transmitter, receiver, and object locations, but coated objects can have nearly length-independent reflection probability.Appropriate placement can allow small coated objects to achieve high reflector likelihood.
- Impact of Size on Performance of LIS Systems: Relay-station average SNR grows linearly with N, whereas large and small intelligent surfaces show quadratic growth with N.The surveyed comparison attributes the LIS scaling to each element acting as a separate reflecting mirror.
- Impact of Size on Performance of LIS Systems: LIS can significantly double transmission rate for specific N values at 2.6 GHz and 28 GHz.The comparison reports this result for LIS relative to relay stations and small intelligent surfaces.
VI. THE POTENTIAL OF POSITIONING AND COVERAGE IN LIS SYSTEMS
LIS positioning studies analyze RSS-based localization, centralized versus distributed deployment, and spherical versus planar surfaces. The surveyed results associate distributed and spherical designs with improved positioning bounds or coverage, while identifying deployment and hardware trade-offs.
- Positioning in Centralized and Distributed LIS Systems: LIS positioning research uses RSS and CRLB analysis to assess UE localization in indoor and outdoor environments.The motivation includes high-accuracy positioning with mmWave and THz technologies.
- Positioning in Centralized and Distributed LIS Systems: Distributed LIS has lower x- and y-dimension CRLBs and significantly better coverage probability than centralized LIS with the same total surface area.Centralized LIS treats each surface as one unit, whereas distributed LIS divides it into smaller independent intelligent units.
- Positioning in Centralized and Distributed LIS Systems: Distributed deployment enables flexible unit replacement and can minimize hardware impairments, but requires specialized calibration hardware and cooperation among subunits.The cited discussion also notes increased complexity and feedback overhead.
- Positioning in Centralized and Distributed LIS Systems: LIS-assisted mmWave positioning is evaluated against conventional mmWave positioning using CRLB performance bounds.The comparison leverages the LIS phase and number of elements to support positioning accuracy.
- Positioning Using Spherical LIS: Spherical LIS has a smaller CRLB than planar LIS, making it more accurate for UE positioning.The spherical surface can combine reflecting and relaying functions, including support for blocked UEs.
A. Realistic Optimization Frameworks
The paper identifies practical optimization directions for LIS-assisted networks, addressing unrealistic assumptions, unexamined objectives, computational complexity, hybrid communication, health constraints, and localization. It also surveys LIS operation, optimization, performance analysis, and open research challenges.
- Realistic system assumptions: Future LIS optimization should replace assumptions such as perfect CSI, negligible internal losses, far-field radiation, and ideal beamforming with realistic models.The paper calls for examining IRS reliability under practical conditions.
- Unaddressed optimization objectives: Beyond energy efficiency, throughput, and SINR, proposed directions include reliability-constrained power minimization for ultra-reliable low-latency communication.The formulation can account for queue-length exceedance events and unknown CSI through distributed federated learning.
- Energy sustainability: Energy harvesting could make LIS systems more sustainable by converting incident radio signals into electrical energy, although RF sources provide low incident power.Available harvested power depends mainly on transmitted frequency and antenna gain, among other factors.
- Algorithmic complexity: Alternating optimization can be computationally costly and may jeopardize joint optimality, motivating data-driven methods based on deep, transfer, and reinforcement learning.These approaches are proposed to reduce the complexity of analyzing and designing IRS systems.
- System integration: Future systems may combine RIS with full-duplex communication, cooperative NOMA, and hybrid RF-VLC architectures to address power allocation, beamforming, and reliable communication needs.The surveyed directions also include tunable materials for adapting phase shifts and supporting broad frequency response.
- Health and localization: Health-aware LIS design should optimize data rate subject to exposure constraints, while localization requires jointly designing beamformers and phase shifters and solving reflector-placement problems.Optimal deployment is described as an open inverse problem of channel modeling.