arxiv: 2604.19056 · v1 · submitted 2026-04-21 · 📡 eess.SP

Recognition: unknown

QoS-Constrained Scheduling in Multi-Cell Multi-User MIMO Networks

Tenghao Cai , Lei Li , Tsung-Hui Chang

Authors on Pith no claims yet

Pith reviewed 2026-05-10 02:36 UTC · model grok-4.3

classification 📡 eess.SP

keywords QoS-constrained schedulingmulti-cell MU-MIMOjoint transmissioneigen-based zero-forcingmassive MIMO asymptoticsblock coordinate descentthroughput maximizationcarrier aggregation

0 comments

The pith

A tractable surrogate using eigen-based zero-forcing beamforming lets schedulers maximize throughput in multi-cell MU-MIMO networks while meeting user rate constraints.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The authors want to allocate users to carriers and base stations, and choose beamforming vectors, so that the total network throughput is as large as possible while every user still receives at least a prescribed minimum rate. The difficulty is that the choice of which users to schedule on which resources is tightly coupled to the exact beamforming vectors that will be used, producing a non-convex problem that cannot be solved at scale. By replacing the true beamforming with an eigen-based zero-forcing design whose performance can be expressed through massive-MIMO asymptotic formulas, they obtain a surrogate problem whose variables separate into blocks that a penalty-based block coordinate descent method can optimize efficiently. When the approximation holds, the resulting schedules satisfy the rate constraints and deliver higher aggregate throughput than earlier schemes that either ignore joint transmission or use simpler heuristics.

Core claim

The original non-convex joint scheduling and beamforming problem can be replaced by a separable surrogate that employs eigen-based zero-forcing beamforming together with massive MIMO asymptotic rate expressions; a penalty-based block coordinate descent algorithm then solves the surrogate to produce feasible schedules and beamformers that meet the QoS rate constraints for both joint-transmission and non-joint-transmission users.

What carries the argument

The eigen-based zero-forcing beamforming combined with massive MIMO asymptotics, which acts as a surrogate that decouples scheduling variables from the exact beamforming vectors and yields a block-separable optimization.

If this is right

The scheduler satisfies the prescribed minimum-rate QoS constraints for every user.
It produces higher total throughput than existing multi-cell scheduling methods that do not use the same approximation.
The block-separable structure permits efficient iterative solution even when carrier aggregation and joint transmission are both present.
The penalty formulation guarantees that the returned solution respects the rate constraints at convergence.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same surrogate technique could be reused for related resource-allocation tasks such as power control or user association in the same network model.
Because the blocks separate by carrier and by base station, the algorithm lends itself to distributed implementation across cooperating cells.
If the asymptotic expressions are replaced by more accurate finite-antenna bounds, the same algorithmic skeleton might still apply with only modest extra computation.

Load-bearing premise

The eigen-based zero-forcing beamforming and massive MIMO rate approximations remain accurate enough that the scheduling decisions they produce are close to those of the true joint problem.

What would settle it

A direct comparison, on the same network instances, between the throughput of the proposed scheduler and the throughput obtained by solving the original non-convex joint problem to global optimality with a much more expensive method.

Figures

Figures reproduced from arXiv: 2604.19056 by Lei Li, Tenghao Cai, Tsung-Hui Chang.

**Figure 1.** Figure 1: System model. structure. Then, we introduce a penalty-based block coordinate descent (BCD) algorithm to handle the surrogate problem efficiently. Extensive results demonstrate that the proposed scheduling algorithm not only ensures QoS satisfaction but also achieves significant throughput gains over existing schedulers. II. SYSTEM MODEL & PROBLEM FORMULATION A. System Model As shown in [PITH_FULL_IMAGE:f… view at source ↗

**Figure 2.** Figure 2: Convergence of the proposed scheduling algorithm. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Figure 3: ESR versus the association threshold of JT-UEs, [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

**Figure 4.** Figure 4: ESR (bar charts) and Sat (dashed-line-connected symbols) achieved by different schemes versus different numbers of UEs K, Nt = 64. the highest ESR while maintaining 100% Sat. In contrast, both mSHS and SUS exhibit significant performance loss. The mSHS, while improving upon SHS by introducing QoS-aware weighting, still suffers from low ESR and Sat. This is largely due to its conservative single-UE-per-RBG … view at source ↗

read the original abstract

In 5G and beyond networks, efficient scheduling is essential to exploit the gains of multi-user MIMO (MU-MIMO) equipped with carrier aggregation and joint transmission (JT). However, cross-cell and cross-carrier scheduling under QoS constraints is challenging due to the strong coupling across users, base stations, and carriers. In this work, we address this problem in multi-cell MU-MIMO networks to maximize system throughput for both JT and non-JT users under rate constraints. The optimization is highly complex as scheduling variables and beamforming (BF) vectors are intertwined. To tackle it, we propose an approximate but tractable surrogate by leveraging the eigen-based zero-forcing BF and massive MIMO asymptotics. The reformulated problem has a separable structure and is amenable to efficient solutions by a penalty-based block coordinate descent method. Simulations demonstrate that the proposed scheduler not only meets the QoS requirements well but also achieves remarkable throughput gains over existing schemes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

A practical BCD solver for multi-cell MU-MIMO QoS scheduling that uses eigen-ZF and large-M asymptotics, but without error bounds on the surrogate for finite antennas.

read the letter

The paper's main contribution is a penalty-based block coordinate descent algorithm that turns the joint scheduling and beamforming problem into a more separable form by fixing eigen-based zero-forcing precoders and replacing exact rates with massive-MIMO asymptotic expressions. This lets them handle both joint-transmission and non-JT users under minimum-rate constraints across cells and carriers. The approach is a direct extension of existing MU-MIMO optimization tricks, and the simulations indicate it meets the QoS targets while delivering throughput gains over the baselines they compare against. That part is useful for people who need a working scheduler rather than a new theoretical framework. The soft spot is exactly the one the stress-test note flags: they do not bound or measure how much the asymptotic surrogate deviates from the true finite-M, finite-K SINR values once the schedule is fixed. Without that check, or even a small-instance comparison against a high-fidelity solver, it is possible the reported feasibility and gains are optimistic when QoS constraints are active. The math itself looks standard and the citation pattern is appropriate, but the validation gap is real. This work is for wireless systems researchers who build or tune schedulers in 5G-style networks. It shows honest engagement with the practical constraints and uses reproducible optimization steps, so it deserves a serious referee who can press on the approximation quality and ask for tighter validation experiments.

Referee Report

3 major / 2 minor

Summary. The paper addresses QoS-constrained joint scheduling and beamforming in multi-cell MU-MIMO networks supporting carrier aggregation and joint transmission (JT). It replaces the original non-convex problem with a surrogate that employs eigen-based zero-forcing precoders and massive-MIMO asymptotic rate expressions, yielding a separable structure solved by a penalty-based block coordinate descent algorithm. Simulations are used to claim that the resulting schedules satisfy rate constraints while delivering substantial throughput gains relative to existing schemes.

Significance. If the surrogate is shown to be sufficiently accurate, the approach would provide a computationally practical method for a relevant 5G scheduling task. The separable reformulation and use of standard asymptotic tools are constructive, but the central throughput and feasibility claims rest on unquantified approximation quality for finite antenna counts, limiting the result's immediate impact.

major comments (3)

[Section 3 (System Model and Surrogate Reformulation)] The surrogate replaces exact SINR expressions with eigen-ZF precoders and large-M asymptotic rates (abstract and Section 3). No analytical bound or error analysis is supplied for the rate approximation under finite M, K, and multi-cell JT conditions; because QoS constraints are enforced only on the surrogate rates, any systematic optimism would directly undermine the reported feasibility and gains.
[Section 4 (Algorithm)] The penalty-based BCD solver is applied exclusively to the surrogate (Section 4). No cross-validation is performed by comparing surrogate solutions against a high-fidelity or globally optimal solution of the original non-convex problem on small instances; without this, it is impossible to confirm that the observed QoS compliance and throughput improvements are not artifacts of the approximation.
[Section 5 (Numerical Results)] Simulation results (Section 5) report that QoS requirements are met and throughput gains are 'remarkable,' yet the figures and tables contain no comparison of surrogate-predicted rates versus actually realized rates (computed with the obtained schedules and finite-M beamformers). This missing check is load-bearing for the central claim.

minor comments (2)

[Section 3] Clarify the precise definition of the eigen-ZF precoder (e.g., which eigenvectors are retained) and the exact form of the asymptotic rate formula used in the surrogate.
[Section 5] Add a table summarizing simulation parameters (M, K, number of carriers, JT vs. non-JT fractions) to improve reproducibility.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We address each major comment below and will incorporate revisions to strengthen the validation of the surrogate approach.

read point-by-point responses

Referee: [Section 3 (System Model and Surrogate Reformulation)] The surrogate replaces exact SINR expressions with eigen-ZF precoders and large-M asymptotic rates (abstract and Section 3). No analytical bound or error analysis is supplied for the rate approximation under finite M, K, and multi-cell JT conditions; because QoS constraints are enforced only on the surrogate rates, any systematic optimism would directly undermine the reported feasibility and gains.

Authors: We acknowledge that a rigorous analytical error bound for the asymptotic approximation under finite M, K and multi-cell JT would be desirable. The massive-MIMO rate expressions we employ are standard in the literature and converge to the exact values as M grows large; our simulations use M=64, which is representative of practical massive-MIMO deployments. In the revision we will add a dedicated numerical study quantifying the gap between the surrogate rates and the exact finite-M rates (computed with the eigen-ZF precoders) for the parameter sets used in the paper. This will explicitly assess any systematic optimism and its impact on the reported QoS feasibility. revision: partial
Referee: [Section 4 (Algorithm)] The penalty-based BCD solver is applied exclusively to the surrogate (Section 4). No cross-validation is performed by comparing surrogate solutions against a high-fidelity or globally optimal solution of the original non-convex problem on small instances; without this, it is impossible to confirm that the observed QoS compliance and throughput improvements are not artifacts of the approximation.

Authors: We agree that explicit cross-validation on small instances would increase confidence. The original joint scheduling-and-beamforming problem is combinatorial and non-convex, rendering global optimality intractable even for modest sizes. For the revision we will add a small-scale validation (e.g., 2-cell, reduced user set) that solves the original problem via exhaustive enumeration of feasible schedules combined with convex beamforming subproblems (or semidefinite relaxation) and compares the resulting throughput and feasibility against the surrogate solution. This will demonstrate that the surrogate does not produce artifacts. revision: yes
Referee: [Section 5 (Numerical Results)] Simulation results (Section 5) report that QoS requirements are met and throughput gains are 'remarkable,' yet the figures and tables contain no comparison of surrogate-predicted rates versus actually realized rates (computed with the obtained schedules and finite-M beamformers). This missing check is load-bearing for the central claim.

Authors: We thank the referee for identifying this important omission. In the revised manuscript we will augment Section 5 with new results that, for every reported schedule, compute the exact finite-M SINR values using the eigen-ZF beamformers and compare them directly with the surrogate rates. We will add tables and/or figures showing both the surrogate-predicted and realized rates, together with the fraction of users that satisfy the QoS constraints under the exact model. Any observed discrepancies will be discussed, thereby confirming that the reported QoS compliance and throughput gains are not artifacts of the approximation. revision: yes

Circularity Check

0 steps flagged

No circularity: surrogate derived from standard asymptotics, solved independently, evaluated by simulation

full rationale

The derivation proceeds from the original non-convex joint scheduling-and-beamforming problem to an approximate surrogate obtained by substituting eigen-based ZF precoders and large-M asymptotic rate formulas (standard massive-MIMO results, not self-derived). The surrogate is then solved via penalty-based block coordinate descent, an off-the-shelf method. Throughput gains and QoS satisfaction are reported from direct numerical comparison of the resulting scheduler against baseline schemes on finite-M instances. None of these steps reduces the claimed performance to a fitted parameter, a self-citation chain, or a definitional tautology; the central claims rest on external simulation benchmarks rather than internal re-labeling of inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on the validity of eigen-based ZF as a proxy for optimal beamforming and the accuracy of massive MIMO asymptotic approximations in finite-antenna multi-cell settings; no free parameters or invented entities are mentioned in the abstract.

axioms (1)

domain assumption Eigen-based zero-forcing beamforming combined with massive MIMO asymptotics yields a tractable and sufficiently accurate surrogate for the original joint scheduling-beamforming problem.
Invoked to reformulate the highly complex optimization into a separable problem.

pith-pipeline@v0.9.0 · 5462 in / 1272 out tokens · 22308 ms · 2026-05-10T02:36:14.126192+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

16 extracted references · 1 canonical work pages

[1]

Optimizing downlink resource allocation in multiuser MIMO networks via fractional programming and the hungarian algorithm,

A. A. Khan, R. S. Adve, and W. Yu, “Optimizing downlink resource allocation in multiuser MIMO networks via fractional programming and the hungarian algorithm,”IEEE Trans. Wireless Commun., vol. 19, no. 8, pp. 5162–5175, 2020

2020
[2]

M 3: A sub-millisecond scheduler for multi-cell MIMO networks under C-RAN architecture,

Y . Chen, Y . T. Hou, W. Lou, J. H. Reed, and S. Kompella, “M 3: A sub-millisecond scheduler for multi-cell MIMO networks under C-RAN architecture,” inProc. IEEE INFOCOM, London, United Kingdom, May 2022, pp. 130–139

2022
[3]

mCore: Achieving Sub- millisecond Scheduling for 5G MU-MIMO Systems,

Y . Chen, Y . Wu, Y . T. Hou, and W. Lou, “mCore: Achieving Sub- millisecond Scheduling for 5G MU-MIMO Systems,” inProc. IEEE INFOCOM, Vancouver, BC, Canada, May 2021, pp. 1–10

2021
[4]

Coordinated scheduling and beamforming for multicell spectrum sharing networks using branch & bound,

L. Yu, E. Karipidis, and E. G. Larsson, “Coordinated scheduling and beamforming for multicell spectrum sharing networks using branch & bound,” inProc. EUSIPCO, Bucharest, Romania, Oct. 2012, pp. 819– 823

2012
[5]

Cross-layer optimization: Joint user scheduling and beamforming design with qos support in joint transmission networks,

S. He, Z. An, J. Zhu, M. Zhang, Y . Huang, and Y . Zhang, “Cross-layer optimization: Joint user scheduling and beamforming design with qos support in joint transmission networks,”IEEE Trans. Commun., vol. 71, no. 2, pp. 792–807, 2022

2022
[6]

Multicell coordinated scheduling with multiuser zero-forcing beamforming,

M. Li, I. B. Collings, S. V . Hanly, C. Liu, and P. Whiting, “Multicell coordinated scheduling with multiuser zero-forcing beamforming,”IEEE Trans. Wireless Commun., vol. 15, no. 2, pp. 827–842, 2015

2015
[7]

Communication- efficient decentralized linear precoding for massive MU-MIMO sys- tems,

X. Zhao, M. Li, Y . Liu, T.-H. Chang, and Q. Shi, “Communication- efficient decentralized linear precoding for massive MU-MIMO sys- tems,”IEEE Trans. Signal Process., vol. 71, pp. 4045–4059, 2023

2023
[8]

On the optimality of multiantenna broad- cast scheduling using zero-forcing beamforming,

T. Yoo and A. Goldsmith, “On the optimality of multiantenna broad- cast scheduling using zero-forcing beamforming,”IEEE J. Sel. Areas Commun., vol. 24, no. 3, pp. 528–541, 2006

2006
[9]

Throughput maximization in clustered cellular networks by using joint resource scheduling and fractional frequency reuse-aided coordinated multipoint,

M. Idrees, X. Qi, A. Zaib, A. Tariq, I. Ullah, Z. Mahmood, and S. Khattak, “Throughput maximization in clustered cellular networks by using joint resource scheduling and fractional frequency reuse-aided coordinated multipoint,”Arab. J. Sci. Eng., pp. 1–15, 2022

2022
[10]

Eigen-based transceivers for the MIMO broadcast channel with semi-orthogonal user selection,

L. Sun and M. R. McKay, “Eigen-based transceivers for the MIMO broadcast channel with semi-orthogonal user selection,”IEEE Trans. Signal Process., vol. 58, no. 10, pp. 5246–5261, 2010

2010
[11]

Decentralizing Coherent Joint Transmission Precoding via Fast ADMM with Deterministic Equivalents,

X. Bian, Y . Liu, Y . Xu, T. Hou, W. Wang, Y . Mao, and J. Zhang, “Decentralizing Coherent Joint Transmission Precoding via Fast ADMM with Deterministic Equivalents,”arXiv preprint arXiv:2403.19127, 2024

work page arXiv 2024
[12]

High-dimensional probability,

R. Vershynin, “High-dimensional probability,” 2009

2009
[13]

Nocedal and S

J. Nocedal and S. J. Wright,Numerical optimization. Springer, 1999

1999
[14]

A framework for one-bit and constant-envelope precoding over multiuser massive MISO channels,

M. Shao, Q. Li, W.-K. Ma, and A. M.-C. So, “A framework for one-bit and constant-envelope precoding over multiuser massive MISO channels,”IEEE Trans. Signal Process., vol. 67, no. 20, pp. 5309–5324, 2019

2019
[15]

QuaDRiGa: A 3- D multi-cell channel model with time evolution for enabling virtual field trials,

S. Jaeckel, L. Raschkowski, K. B ¨orner, and L. Thiele, “QuaDRiGa: A 3- D multi-cell channel model with time evolution for enabling virtual field trials,”IEEE Trans. Antennas Propag., vol. 62, no. 6, pp. 3242–3256, 2014

2014
[16]

Analysis of Non-Coherent Joint-Transmission cooperation in Heterogeneous Cellu- lar Networks,

R. Tanbourgi, S. Singh, J. G. Andrews, and F. K. Jondral, “Analysis of Non-Coherent Joint-Transmission cooperation in Heterogeneous Cellu- lar Networks,” inProc. IEEE ICC, Sydney, NSW, Australia, Jun. 2014, pp. 5160–5165

2014