Source-linked AI summary
RIS-Assisted Reliability Maximization for URLLC in Upper Mid-Band 6G Networks
Muddasir Rahim, Irfan Azam, Soumaya Cherkaoui
TL;DR
FR3 RIS-assisted URLLC faces reliability challenges, while existing approaches often separate power allocation from user association and Shannon formulations do not capture short-packet requirements. This paper jointly optimizes power and UE–RIS association under finite-blocklength constraints using a multiphase SCA and matching framework; simulations report performance close to exhaustive search and better than greedy and random baselines.
Problem
RIS-assisted FR3 URLLC remains relatively underexplored, and existing RIS work often optimizes power or association independently while Shannon formulations are inadequate for short-packet reliability requirements.
Source: §I · pp. 1–2Method
The paper jointly optimizes transmit power and UE–RIS association under finite-blocklength constraints using SCA-based power optimization and matching-theory association.
Source: §I · p. 2Results
The matching-based scheme outperforms SCA-GS and SCA-RS and achieves performance close to SCA-ES.
Source: §IV · p. 5Takeaways & Limitations
The results highlight the importance of jointly optimizing power allocation and RIS association in FR3 networks.
Source: §V · p. 6
Abstract
from arXiv · showhide
In sixth-generation (6G) networks, supporting ultra-reliable low-latency communication (URLLC) services imposes stringent requirements on reliability, latency, and efficient resource utilization. The upper mid-band (FR3) spectrum has emerged as a promising candidate due to its favorable trade-off between coverage and bandwidth availability. However, achieving high reliability in dense deployments remains challenging due to interference and propagation impairments. To address these challenges, this paper investigates a reconfigurable intelligent surface (RIS)-assisted network operating in the FR3 band. We formulate a joint transmit power allocation and user equipment (UE)-RIS association problem to maximize network reliability under finite blocklength constraints. The resulting problem is nonconvex and combinatorial due to interference coupling and binary association variables. To solve this problem efficiently, we propose a multiphase resource allocation framework that integrates a successive convex approximation (SCA)-based power optimization scheme with a matching theory-based UE-RIS association algorithm. Simulation results show that the proposed scheme outperforms greedy and random baselines and achieves near-optimal performance close to the exhaustive search scheme.
I. INTRODUCTION
FR3 offers a bandwidth–coverage trade-off for 6G, but RIS-assisted URLLC reliability in this band remains underexplored. The paper jointly optimizes transmit power and UE–RIS association under finite-blocklength constraints and proposes a multiphase solution.
I. INTRODUCTION
- FR3 is presented as a 6G candidate offering a trade-off between bandwidth availability and coverage, while path loss and LoS blockages challenge reliability in dense deployments.Source: §I · p. 1
- Prior RIS work is concentrated on FR2 and single-tier terrestrial settings; multi-tier FR3 operation, including aerial and terrestrial RISs, has received comparatively little attention.The introduction also notes that FR3 MIMO studies remain limited and reviews work on UAV and aerial-terrestrial RIS networks.Source: §I · pp. 1–2
- Existing RIS studies often optimize power allocation or user association independently, while Shannon-capacity formulations do not capture short-packet URLLC requirements adequately.Source: §I · p. 2
- The proposed FR3 RIS model incorporates finite blocklength and jointly optimizes transmit power allocation and UE–RIS association to maximize network reliability.Source: §I · p. 2
- A multiphase framework combines SCA-based power optimization with matching-theory-based UE–RIS association to solve the nonconvex, combinatorial problem.The method is presented as enabling efficient and scalable implementation.Source: §I · p. 2
- The matching-based scheme outperforms greedy and random baselines and achieves performance close to exhaustive search.The simulation discussion also reports that it performs close to SCA-ES and consistently outperforms SCA-GS and SCA-RS across the considered power range.Source: §I · p. 2
II. SYSTEM MODEL
The system model considers an FR3 downlink with direct and RIS-reflected paths, finite-blocklength URLLC transmission, and a joint power-allocation and UE–RIS-association objective. The formulation includes power and one-to-one association constraints and is mixed-integer and nonconvex.
II. SYSTEM MODEL
- The network has one multi-antenna AP, UEs, and multiple RISs; UEs can receive direct signals or RIS-reflected signals when the direct path is weak or blocked.The AP–RIS and RIS–UE channels are modeled as quasi-static fading channels in FR3.Source: §II · p. 2
- The model includes direct AP–UE and RIS-assisted channels, with a binary UE–RIS association variable and RIS phase shifts fixed using statistical CSI.The fixed phase design uses large-scale channel parameters and updates on a slower timescale to reduce signaling overhead.Source: 1) Channel Model · p. 2
- AP beamforming vectors are scaled by per-UE transmit powers, whose sum is constrained by the total power budget.Source: 2) Signal Model and Data Rate of URLLC · pp. 2–3
- URLLC rates use a finite-blocklength formula, and UE reliability is defined from the decoding error probability.The passage identifies packet length, error probability, and channel dispersion as terms in the rate formulation.Source: 2) Signal Model and Data Rate of URLLC · p. 3
- The optimization maximizes network sum reliability under finite blocklength while jointly selecting transmit powers and UE–RIS associations.The constraints impose positive powers, a total-power limit, binary assignments, and at most one UE per RIS and one RIS per UE; the resulting problem is mixed-integer and nonconvex.Source: 3) Optimization Problem Formulation · p. 3
III. PROPOSED RESOURCE ALLOCATION FRAMEWORK
The proposed resource-allocation framework solves the NP-hard problem sequentially, optimizing downlink transmit powers before UE–RIS association.
III. PROPOSED RESOURCE ALLOCATION FRAMEWORK
- The AP first optimizes downlink transmit power, then optimizes UE–RIS association in a sequential framework for the NP-hard problem.The passage characterizes this decoupling as enabling efficient resource allocation while enhancing received signal quality and overall FR3 network reliability.Source: §III · p. 3
A. SCA-based Power Optimization
With the RIS association fixed, the section develops an SCA-based power optimization that converts the reliability objective into a convex subproblem at each iteration. The resulting subproblem supports the power-optimization phase of the resource-allocation algorithm.
A. SCA-based Power Optimization
- Given the RIS association matrix, the algorithm optimizes UE downlink transmit powers in Phase 1 using the SCA subproblem.The convex subproblem can be solved with standard convex solvers such as CVX.Source: A. SCA-based Power Optimization · pp. 3–5
- The reliability objective is reformulated by minimizing error probability, then approximated through an auxiliary variable and a fixed channel-dispersion value at each SCA iteration.Because the Q-function is monotonically decreasing, minimizing error probability is equivalent to maximizing the auxiliary objective.Source: A. SCA-based Power Optimization · p. 4
- The power objective remains nonconvex because it is a difference of two concave terms, so SCA linearizes the second term at the current iterate.The first-order Taylor expansion upper-bounds that term and guarantees monotonic improvement in SCA.Source: A. SCA-based Power Optimization · p. 4
- For a given RIS association matrix, the transformed subproblem P1.1 is convex.Substituting the Taylor bound yields a concave objective, and constraints C1 and C2 are convex.Source: A. SCA-based Power Optimization · p. 4
B. Deferred Acceptance Matching-based UE-RIS Association
The UE-RIS association is formulated as a one-to-one matching problem using reliability-based preferences. A UE-proposing deferred-acceptance procedure yields a stable matching with a polynomial proposal bound.
B. Deferred Acceptance Matching-based UE-RIS Association
- The power allocation from the preceding subproblems is used to formulate the UE-RIS association problem.Each UE can be associated with at most one RIS in a transmission slot.Source: B. Deferred Acceptance Matching-based… · p. 5
- UEs rank RISs by achievable reliability, and each RIS retains the proposing UE with the highest reliability.Unmatched UEs propose to their highest-ranked untried RIS; rejected UEs proceed to their next choice.Source: B. Deferred Acceptance Matching-based… · p. 5
- The deferred-acceptance procedure terminates after at most K × L proposals and produces a stable association.Each UE proposes to at most L RISs, and the resulting matching has no blocking pair.Source: B. Deferred Acceptance Matching-based… · p. 5
IV. SIMULATION RESULTS AND DISCUSSION
Across the evaluated power, blocklength, and RIS-element settings, the matching-based scheme stays close to exhaustive search and outperforms greedy and random association baselines. Reliability increases with each of these resource or transmission parameters for the considered schemes.
IV. SIMULATION RESULTS AND DISCUSSION
- The simulations use a 64-antenna AP, 100×100-reflector RISs, a 15 GHz carrier, 200 MHz bandwidth, and a 23 dBm AP power budget.Source: §IV · p. 5
- All benchmarks use the same SCA-based power optimizer; SCA-ES exhaustively searches associations, SCA-GS selects greedily, and SCA-RS assigns them randomly.Each data point averages 10^6 independent channel realizations.Source: §IV · p. 5
- The matching-based scheme stays close to SCA-ES and outperforms SCA-GS and SCA-RS across power, blocklength, and RIS-element settings.The cited figures show reliability versus AP power budget, finite blocklength, and number of RIS elements, respectively.Source: Figure 2 · p. 5, §IV · pp. 5–6, Figure 3 · p. 6
- Reliability rises for all schemes as AP power, finite blocklength, and RIS-element count increase.The paper attributes these trends to stronger received signal and improved SINR, coding gain and reduced decoding error, and passive beamforming gain, respectively.Source: Figure 2 · p. 5, §IV · pp. 5–6, Figure 3 · p. 6
V. CONCLUSIONS
The paper formulates reliability maximization for RIS-assisted FR3 networks under finite blocklength constraints and solves the joint allocation problem with a multiphase framework. Simulations show improved reliability over greedy and random association schemes, while future work targets grant-free massive access.
V. CONCLUSIONS
- The paper formulates joint transmit-power allocation and UE–RIS association to maximize reliability in RIS-assisted FR3 networks under finite blocklength constraints.Source: §V · p. 6
- A multiphase framework combines SCA-based power optimization with matching-theory UE–RIS association for the nonconvex, combinatorial problem.Source: §V · p. 6
- Simulations show improved reliability over greedy and random association schemes; future work will add active-user detection and dynamic resource allocation for grant-free massive access.The proposed extension is aimed at improving scalability and robustness under time-varying network conditions.Source: §V · p. 6