Source-linked AI summary
QoS-Aware Joint Concurrency and HARQ-Limit Design for HARQ-CC-Aided Slow Fluid Antenna Multiple Access
Sixu Han, Kai-Kit Wong, Hanjiang Hong, Haoyu Liang
TL;DR
One-shot sFAMA can lack reliability at high concurrency or stringent spectral efficiency, while HARQ-CC trades longer service for improved reliability. This paper jointly designs user concurrency and HARQ limits using selected-port SIR analysis and a two-stage QoS optimization. In the reported evaluation, adaptive FAS supports six concurrent users at dmax = 4, exceeding the concurrency of the fixed-HARQ FAS and adaptive FPA comparisons.
Problem
One-shot sFAMA may lack sufficient reliability at high user concurrency or stringent spectral-efficiency requirements, while concurrency and HARQ limits trade interference against retransmission-based reliability and service duration.
Source: §I · p. 1Method
The paper characterizes multi-round decoding failure from selected-port SIR distributions and jointly chooses maximum feasible concurrency and the minimum HARQ limit satisfying outage and mean-service constraints.
Source: §I · p. 1Results
At dmax = 4, the adaptive-FAS design maintains U⋆ = 6 over the considered outage range, while fixed-HARQ FAS and FPA achieve smaller concurrency levels.
Source: §V · p. 5Takeaways & Limitations
The reported results show that joint concurrency and HARQ adaptation extends the QoS-feasible concurrency region of FAS.
Source: §VI · p. 5
Abstract
from arXiv · showhide
Hybrid automatic repeat request with Chase combining (HARQ-CC) improves the reliability of slow fluid antenna multiple access (sFAMA), but its transmission limit must be jointly optimized with user concurrency. This paper proposes a quality-of-service (QoS)-aware joint design of concurrency level U and HARQ limit C for downlink HARQ-CC-aided sFAMA. Based on the selected-port signal-to-interference ratio (SIR) distribution, we derive multi-round decoding failure probabilities and develop a two-stage optimization that maximizes feasible concurrency and then selects the minimum C satisfying outage and mean-service constraints. We prove that the first feasible C minimizes service duration for a given U, and that feasible U values form an initial segment, enabling optimal bisection search. Numerical results agree with Monte Carlo simulations under full Jakes spatial covariance. The proposed adaptive FAS design supports six concurrent users, compared with three to five for fixed-HARQ FAS and two for HARQ-adaptive fixed-position-antenna (FPA), demonstrating that joint concurrency and HARQ adaptation expands the QoS-feasible region of FAS.
I. INTRODUCTION
The introduction frames HARQ-CC and user concurrency as coupled QoS design choices for sFAMA: more users increase interference, while more retransmissions improve reliability but lengthen service. It identifies a scheduler-admission gap and presents a joint design with structural properties and simulation validation.
I. INTRODUCTION
- FAS selects the most favorable port, and sFAMA chooses the port maximizing average SIR with slower updates than symbol-level fFAMA switching.FAMA allows multiple downlink streams to share a time–frequency resource while mitigating multiuser interference.Source: §I · p. 1
- One-shot sFAMA may provide insufficient reliability at high concurrency or stringent spectral efficiency, motivating HARQ-CC retransmissions that combine received replicas.More active users increase interference, whereas a larger HARQ limit improves reliability at the cost of longer service duration.Source: §I · p. 1
- Prior HARQ-sFAMA studies model queue–HARQ coupling or exogenous interferer activity, but do not directly solve the scheduler’s admitted-user and minimum-HARQ-limit problem.The paper instead considers full-buffer scheduling, with each departing packet replaced and each user facing U −1 active interferers per round.Source: §I · p. 1
- The proposed design jointly optimizes concurrent users and HARQ limit under packet-outage and mean-service-duration constraints, characterizing QoS through selected-port SIR distributions and Stieltjes convolution.The paper establishes that the first reliability-feasible HARQ limit minimizes mean service duration, while feasible concurrency levels form an initial segment that permits bisection.Source: §I · p. 1
- The design is validated by Monte Carlo simulations using full Jakes spatial covariance and compared with fixed-HARQ FAS and HARQ-adaptive FPA benchmarks.The supplied passage begins to report that joint adaptation extends the QoS-feasible concurrency region, but its final clause is truncated.Source: §I · p. 1
A. System Configuration
The system pairs a multi-antenna downlink serving U users with a one-dimensional FAS receiver at each user. Under sFAMA, each receiver selects a port by SIR, with port correlation modeled spatially and fading blocks independent across HARQ rounds.
A. System Configuration
- The base station uses Nt = U separated transmit antennas to serve U users simultaneously on the same time–frequency resource, one stream per user.Each user has a one-dimensional FAS with K uniformly spaced ports over an aperture of length Wλc and one RF chain accessing one port at a time.Source: A. System Configuration · p. 2
- HARQ rounds experience independent quasistatic fading blocks, while spatial correlation among a user’s K ports within a block follows the Jakes model.The block-correlation approximation groups ports into disjoint blocks, with within-block correlation and independence across blocks.Source: A. System Configuration · p. 2
- Under sFAMA, each user selects an SIR-maximizing port in each HARQ round, producing a selected-port SIR for that round.The port channel vector is modeled using Jakes spatial correlation; the supplied text also states that the noise is neglected under an interference-limited assumption.Source: A. System Configuration · p. 2
- For fixed U, the selected-port SIRs are identically distributed and independent across HARQ rounds, with their common CDF specified under the adopted spatial block-correlation model.The CDF uses U −1 interfering streams; its dependence on K and other fixed parameters is suppressed in the notation.Source: A. System Configuration · p. 2
B. HARQ-CC Reliability and Service Metrics
The reliability and service metrics follow from accumulating independent selected-port SIRs across HARQ rounds and testing the accumulated SIR against a decoding threshold. The resulting outage and mean-service quantities characterize packet performance under a transmission limit.
B. HARQ-CC Reliability and Service Metrics
- Chase combining accumulates selected-port SIR observations across transmissions, and decoding succeeds when the accumulated SIR reaches the threshold set by the initial spectral efficiency.The analysis assumes ideal outage-based decoding, perfect error detection, and instantaneous error-free ACK/NACK feedback.Source: B. HARQ-CC Reliability and Service… · pp. 2–3
- Independence across HARQ rounds yields a Stieltjes-convolution recursion for the accumulated-SIR distribution, which determines decoding-failure probabilities after each round.The per-round SIR distribution is continuous, and the accumulated distribution is likewise continuous.Source: B. HARQ-CC Reliability and Service… · p. 3
- The transmission limit C bounds the number of transmissions per packet, while the actual occupied transmissions are random and determine packet outage and average HARQ service duration.Service duration counts HARQ service time only and excludes queueing delay.Source: B. HARQ-CC Reliability and Service… · p. 3
- Per-user and aggregate resource-normalized goodputs are defined as supplementary resource-efficiency metrics alongside the QoS measures.The supplied passages identify the goodput measures but do not include their formulas.Source: B. HARQ-CC Reliability and Service… · p. 3
III. OPTIMIZATION VARIABLES AND PROBLEM FORMULATION
The design maximizes the number of concurrent users satisfying outage and mean-service constraints, then chooses the smallest HARQ limit among policies supporting that concurrency.
III. OPTIMIZATION VARIABLES AND PROBLEM FORMULATION
- U sets scheduled spatial reuse and the deterministic interferer count U−1, while C sets each packet’s maximum HARQ protection.K is a receiver-hardware parameter and R0 is fixed by the payload and modulation-and-coding mode.Source: §III · p. 3
- The algorithm tests each candidate U using the first c with qi(U) ≤ ϵ, then checks whether its mean service duration meets Dmax.It uses bisection over U and returns the selected pair (U⋆, C⋆), or infeasible.Source: §III · p. 3
- K and R0 may be swept externally for sensitivity analysis, but each reported optimizer must use one fixed (K, R0) pair.Source: §III · p. 3
- The design applies a target packet outage probability ϵ and maximum average service duration Dmax as its QoS constraints.Source: §III · p. 3
- The objective first maximizes concurrently served users and then selects the smallest HARQ limit among policies supporting that concurrency.Source: §III · p. 3
- Aggregate resource-normalized goodput GΣ is supplementary and does not replace the per-user reliability and latency requirements.Source: §III · p. 3
IV. STRUCTURAL REDUCTION OF THE TWO-STAGE DESIGN
For fixed K and R0, the structural results reduce the joint design to finding the largest feasible concurrency and using its first reliability-satisfying HARQ limit.
IV. STRUCTURAL REDUCTION OF THE TWO-STAGE DESIGN
- For fixed K and R0, U is feasible exactly when a finite first reliability-satisfying limit Cϵ(U) exists and its mean service duration is at most Dmax.Source: §IV · p. 4
- Feasible concurrency levels form an initial segment: feasibility at U2 implies feasibility at every smaller admissible U1.Source: §IV · p. 4
- The first HARQ limit meeting the reliability target minimizes mean service duration among all reliability-feasible limits for that U.Source: §IV · p. 4
- These properties eliminate C from the outer search and allow globally optimal bisection over U, reducing feasibility evaluations from O(Umax) to O(log Umax).The second-stage choice is C⋆ = Cϵ(U⋆).Source: §IV · p. 4
V. NUMERICAL RESULTS
The numerical results validate the outage analysis against full-Jakes Monte Carlo simulations and show that jointly adapting concurrency and HARQ limit yields QoS-feasible designs. For ϵ = 10−3 and dmax = 4, the selected design supports six users with HARQ limit seven.
V. NUMERICAL RESULTS
- Analytical outage results closely match full-Jakes Monte Carlo simulations; at fixed C, outage rises with U, while larger C improves reliability.FAS consistently outperforms FPA under identical (U, C) configurations.Source: Figure 1 · p. 4, §V · p. 4
- Increasing C lowers outage but increases transmission occupancy, while larger U increases interference and service demand.The reliability–service tradeoff is evaluated against both outage and mean-service constraints.Source: Figure 2 · p. 4, §V · p. 4
- The evaluation varies ϵ from 10−4 to 10−1 and dmax from 1 to 5, using full-Jakes Monte Carlo benchmarks with 2×108 packet realizations.The baseline setup uses 7 GHz, a 0.15 m FAS aperture, 48 ports, and Cmax = 12.Source: §V · p. 4
- At ϵ = 10−3 and dmax = 4, the design selects (U ⋆, C⋆) = (6, 7).It first maximizes feasible concurrency, then selects the minimum HARQ limit.Source: §V · p. 4, Figure 2 · p. 4
- The adaptive-FAS scheme maintains U ⋆= 6 across the considered outage range, while fixed-HARQ FAS and adaptive FPA achieve smaller concurrency levels.Fig. 3a compares these designs at dmax = 4.Source: Figure 3 · p. 5, §V · p. 5
VI. CONCLUSION
The paper jointly designs concurrency and HARQ limits for downlink HARQ-CC-aided sFAMA under outage and mean-service constraints. Selected-port SIR distributions and structural properties support the design, and full-Jakes simulations validate its analysis and show higher QoS-constrained concurrency through adaptation.
VI. CONCLUSION
- The joint design maximizes feasible concurrency under outage and mean-service constraints, then selects the minimum feasible HARQ limit.Selected-port SIR distributions yield multi-round failure probabilities and QoS metrics via Stieltjes convolution.Source: §VI · p. 5
- Monotonicity enables a two-stage global design: bisection finds maximum feasible concurrency before choosing the minimum feasible HARQ limit.Full-Jakes simulations validate the analysis and show that adaptation converts FAS reliability gains into higher QoS-constrained concurrency.Source: §VI · p. 5