Source-linked AI summary
HAPS-RIS or HAPS-Relay: Which Outperforms Under Impairments with NOMA in 6G NTN?
Bilal Karaman, Faicel Khennoufa, Ilhan Basturk, Metin Ozturk, Ferdi Kara, Sezai Taskin, Halim Yanikomeroglu
TL;DR
The paper addresses the limited comparison of HAPS-RIS and HAPS-RS under practical impairments and multiple-access schemes. It evaluates the architectures under OMA and NOMA with HWI and imperfect CSI, finding that HAPS-RIS has higher sum-rate and energy efficiency under non-ideal conditions. It also examines how RIS element allocation and user spacing relate to NOMA performance.
Problem
Comparisons of HAPS-RIS and HAPS-RS under realistic impairments and advanced access schemes are limited, as is evidence on NOMA element allocation and user spacing.
Source: §I · p. 2Method
The paper evaluates HAPS-RIS and HAPS-RS under OMA and NOMA while accounting for HWI and imperfect CSI.
Source: §V · p. 6Results
Under non-ideal conditions, HAPS-RIS achieves superior sum-rate and energy efficiency to HAPS-RS, while HAPS-RS is advantageous for sum rate in the low-power regime.
Source: §IV · p. 5, §V · p. 6Takeaways & Limitations
RIS element allocation and user spatial separation affect NOMA performance, with appropriate allocation and greater separation increasing channel disparity.
Source: §IV · p. 6, §V · p. 6
Abstract
from arXiv · showhide
This paper investigates the performance of high-altitude platform station (HAPS)-assisted communication systems employing either reconfigurable intelligent surfaces (RIS) or relay stations (RS) under non-orthogonal multiple access (NOMA) scheme. Practical system impairments, including hardware impairments (HWI) and imperfect channel state information (CSI), are explicitly considered. The results show that HAPS-RIS outperforms HAPS-RS in terms of both sum-rate and energy efficiency under non-ideal conditions due to its passive nature, which avoids noise amplification. Furthermore, it is demonstrated that RIS element allocation and user spatial distribution significantly impact NOMA performance, where increased user separation and proper allocation enhance channel disparity and improve system efficiency. Despite its higher sensitivity to imperfect CSI, HAPS-RIS can effectively compensate for performance degradation through large-scale RIS element deployment, maintaining a performance advantage over half-duplex RS-based systems. These insights provide useful design guidelines for impairment-aware HAPS-assisted 6G communication systems.
I. Introduction
HAPS can support sustainable, wide-area 6G connectivity, and RIS offers a low-power alternative to active relay payloads. The paper addresses the limited evidence on comparing HAPS-RIS and HAPS-RS under practical impairments and NOMA, including how RIS allocation and user spacing affect performance.
I. Introduction
- HAPS offer a coverage, latency, and deployment trade-off for 6G, while RIS enables low-power signal reflection without dedicated RF chains.HAPS operate around 20 km and can support emergency, temporary, and rural connectivity; active HAPS-RS typically consume about 1 kW.Source: §I · p. 1
- A comprehensive comparison of HAPS-RIS and HAPS-RS under realistic impairments and advanced access schemes remains lacking, as do analyses of NOMA element allocation and user spacing.The paper also identifies HWI and imperfect CSI as insufficiently investigated in these systems.Source: §I · p. 2
- The paper develops a unified framework comparing HAPS-RIS and HAPS-RS under OMA and NOMA while accounting for HWI and imperfect CSI.It characterizes spectral- and energy-efficiency trade-offs and examines RIS allocation and inter-user distance in NOMA.Source: §I · p. 2
- The study evaluates impairment effects, architecture efficiency trade-offs, RIS element allocation, and inter-user distance in NOMA-based HAPS systems.These are identified as contributions alongside a unified HAPS-RIS/HAPS-RS NOMA system model.Source: §I · p. 2
II. System Model
The system model compares HAPS platforms equipped with either an active relay station or a passive RIS, serving multiple ground users through the HAPS. It incorporates hardware impairments and imperfect channel state information.
II. System Model
- The modeled HAPS platform uses either an RS or an N-element RIS to serve multiple ground UEs under NOMA.The system-model figure depicts the RIS- and RS-based architectures.Source: §II · p. 2, Figure 1 · p. 2
- The control station has a high-gain antenna and a line-of-sight link to the HAPS, while no direct control-station-to-user link is assumed.All communications are realized via the HAPS, and each UE has a single antenna.Source: §II · p. 2
- The RS operates as a half-duplex active relay, whereas the RIS passively reflects signals toward users without RF chains.The model accounts for hardware impairments at the control station and UEs and for imperfect channel estimates.Source: §II · p. 2
A. HAPS-RS-assisted Communication
The HAPS-RS model uses superposition coding and allocates power according to ordered effective channel gains. It accounts for imperfect channel estimates, transceiver impairments, and half-duplex relaying in its rate analysis.
A. HAPS-RS-assisted Communication
- The CS uses superposition coding, with power coefficients ordered by effective channel gains and assigned to the users' signals.The coefficients satisfy α1 > α2 > ... > αL and sum to one.Source: A. HAPS-RS-assisted Communication · pp. 2–3
- The model treats HAPS-RS-to-user channel estimation errors as independent zero-mean complex Gaussian variables with variance ϵ.The passage identifies ϵ as the level of CSI imperfection.Source: A. HAPS-RS-assisted Communication · pp. 2–3
- The channel model includes Rician fading, receiver noise, and independent Gaussian distortion noises associated with transceiver impairments.The impairment parameters quantify transmitter- and receiver-side impairment levels.Source: A. HAPS-RS-assisted Communication · p. 3
- For comparison, total transmit power is conserved between HAPS-RS and HAPS-RIS, and the HAPS-RS sum rate accounts for half-duplex relaying.The RS power split is selected through a one-dimensional search that maximizes sum-rate performance.Source: A. HAPS-RS-assisted Communication · p. 3
B. HAPS-RIS-assisted Communication
The HAPS-RIS model allocates RIS elements among users and estimates each user's cascaded end-to-end channel. It derives received-signal and performance expressions while accounting for HWI and imperfect CSI.
B. HAPS-RIS-assisted Communication
- The received-signal model reflects the superposition-coded transmission toward each user through allocated RIS elements, with phase shifts represented by a RIS phase-shift matrix.User i receives N_i elements, with allocations satisfying a total of N elements.Source: B. HAPS-RIS-assisted Communication · p. 3
- The effective cascaded end-to-end channel is estimated because passive RIS elements prevent separate estimation of the CS–RIS and RIS–user channels.The model represents CSI error as a complex Gaussian variable with zero mean and variance proportional to channel power.Source: B. HAPS-RIS-assisted Communication · p. 4
- The section gives the user SINR and achievable sum-rate expressions under HWI and imperfect CSI.Source: B. HAPS-RIS-assisted Communication · p. 4
C. Path Loss Model
The model uses Friis transmission to represent path loss on the CS–HAPS and HAPS–ground-user links. It approximates RIS-element positions because the RIS is small relative to the link distances.
C. Path Loss Model
- Friis transmission models path loss between the CS and HAPS and between the HAPS-mounted RIS and each ground user.Source: C. Path Loss Model · p. 4
- The geometry defines coordinates for the CS, ground user, HAPS, RIS element, and RS to express the large-scale path-loss terms.Source: C. Path Loss Model · p. 4
- Because the RIS is physically small relative to the HAPS altitude and link distances, its element positions are approximated.The text gives RIS dimensions of 5 m × 6 m and CS–HAPS and HAPS–ground-user distances exceeding 20 km.Source: C. Path Loss Model · p. 4
III. Energy Efficiency Analysis
Energy efficiency is defined as achievable sum rate divided by total power consumption. The power model distinguishes RIS element circuitry from RS transmit and payload power.
III. Energy Efficiency Analysis
- Energy efficiency is the ratio of achievable sum rate to total power consumption for either the HAPS-RS or HAPS-RIS system.Source: §III · p. 4
- The RIS power model includes phase-shifting switches and DC-biasing circuits for each element, while the RS model includes RS transmit power and payload-related power.The model also accounts for CS transmit power and each ground user's power consumption.Source: §III · p. 4
IV. Numerical Results and Discussion
Under practical impairments, HAPS-RIS generally outperforms HAPS-RS, while RIS size and user placement shape the trade-offs among sum rate, energy efficiency, and NOMA gain. Element allocation and spatial separation can improve NOMA performance, but their effects depend on the operating conditions.
IV. Numerical Results and Discussion
- Simulations use 105 independent realizations and two-user NOMA; the HAPS-RS payload power assumption is up to 1 kW.The stated user and platform locations define the simulated geometry.Source: §IV · p. 4
- Beyond N ≥15 × 10^3 elements, HAPS-RIS sum rate improves notably under ideal conditions; NOMA outperforms OMA, and HAPS-RIS beats HAPS-RS under non-ideal conditions.Impairments degrade all configurations; the passive RIS avoids the noise amplification associated with active relaying.Source: §IV · pp. 4–5, Figure 2 · p. 5
- HAPS-RIS energy efficiency decreases as RIS size grows, while HAPS-RS energy efficiency remains relatively constant; impairments further reduce efficiency.RIS hardware power consumption grows linearly with element count, while sum-rate gains become marginal beyond a certain point.Source: Figure 3 · p. 5, §IV · p. 5
- HAPS-RS is advantageous at low transmit power, but HAPS-RIS becomes superior at high power; impairments limit gains and cause sum-rate saturation.The comparison uses N = 30 × 10^3 RIS elements for HAPS-RIS.Source: Figure 4 · p. 5, §IV · p. 5
- HAPS-RIS energy efficiency peaks and then declines with transmit power, outperforming HAPS-RS particularly in the moderate-to-high power regime.HWI and imperfect CSI reduce the peak efficiency and shift the optimal operating point.Source: §IV · p. 5, Figure 5 · p. 6
- As the far user moves away, overall sum rate falls while NOMA gain rises, as greater user separation creates stronger channel disparity.In LoS-dominant HAPS-RIS, nearby users tend to have similar channel gains, limiting NOMA's effectiveness.Source: Figure 6 · p. 6, §IV · p. 6
- Allocating more RIS elements to the far user improves sum rate; under HWI and imperfect CSI, the optimal allocation shifts from 2500 to 4500 elements.Equal allocation provides a moderate performance level, while optimized allocation can emulate channel heterogeneity in LoS-dominant HAPS-RIS.Source: Figure 7 · p. 6, §IV · p. 6
- RIS configurations with 5 × 10^3 to 30 × 10^3 elements imply about 4.5–27 m^2 of surface area and 39–234 W of power consumption.The passage compares this power requirement favorably with conventional BS or RS systems.Source: §IV · p. 6
V. Conclusion
Under practical impairments, HAPS-RIS achieves higher sum-rate and energy efficiency than HAPS-RS. RIS allocation and user spatial distribution also shape NOMA performance by affecting channel disparity.
V. Conclusion
- HAPS-RIS achieves superior sum-rate and energy efficiency to HAPS-RS under non-ideal conditions because its passive structure avoids noise and distortion amplification.The conclusion identifies this comparison under practical impairments including HWI and imperfect CSI.Source: §V · p. 6
- RIS element allocation and user spatial distribution critically affect NOMA performance, with appropriate allocation creating channel disparity in LoS-dominant environments.Source: §V · p. 6
- The authors identify adaptive RIS allocation and dynamic user pairing as future work for HAPS-assisted networks.Source: §V · p. 6