pith. sign in

arxiv: 2604.13952 · v1 · submitted 2026-04-15 · 📡 eess.SP

Low-Complexity, Space Splitting-based User Selection in MU-MIMO for Massive Connectivity and AI-Native Traffic

Jo\~ao Paulo S. H. Lima , Marcin L. Filo , Chathura Jayawardena , Konstantinos Nikitopoulos This is my paper

Pith reviewed 2026-05-10 12:27 UTC · model grok-4.3

classification 📡 eess.SP
keywords MU-MIMOuser selectionlow complexitymassive connectivityspectral efficiencyspatial basesAI-native trafficIoT
0
0 comments X

The pith

The Space Splitting-based User Selection algorithm reduces MU-MIMO scheduling complexity by over three orders of magnitude while matching the spectral efficiency of current practical methods.

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

The paper introduces Space Splitting-based User Selection (SS-US) for MU-MIMO systems facing dense, uplink bursty traffic from AI services and massive IoT. Instead of combinatorially searching user subsets, it builds orthonormal spatial bases and matches users independently to directions. This breaks the complexity barrier that grows with more users and layers. Simulations confirm over 1000-fold complexity reduction with comparable spectral efficiency to current methods. Such an approach would allow practical, frequent scheduling in highly dynamic massive connectivity scenarios.

Core claim

The SS-US algorithm departs from subset-based selection by constructing orthonormal spatial bases and independently matching users to spatial directions, thereby reducing computational complexity by over three orders of magnitude while achieving spectral efficiency comparable to state-of-the-art practical baselines in diverse MIMO configurations, channel conditions, and user densities.

What carries the argument

The Space Splitting-based User Selection (SS-US) algorithm, which constructs orthonormal spatial bases and performs independent user-to-direction matching to bypass combinatorial subset optimization.

If this is right

  • Supports low-latency scheduling decisions required for latency-critical AI-native and IoT traffic patterns.
  • Enables MU-MIMO to scale to massive concurrent connectivity without prohibitive processing demands at the base station.
  • Maintains performance parity with existing heuristics in spectral efficiency for uplink-oriented bursty communications.
  • Facilitates practical deployment in dense user scenarios where frequent reselection is needed.

Where Pith is reading between the lines

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

  • SS-US might be combined with predictive models for user arrival to pre-construct bases dynamically.
  • The independent matching could extend to scenarios with partial channel state information or feedback delays.
  • Further gains may appear when applied to extremely large antenna arrays beyond the tested configurations.
  • Real-time hardware tests would be needed to confirm the complexity reduction translates to lower power consumption.

Load-bearing premise

That independently assigning users to pre-built orthonormal spatial bases will control interference sufficiently to keep spectral efficiency close to optimal in all relevant MIMO setups and channel types.

What would settle it

A simulation or field test showing that for certain channel correlation patterns or very high user densities, the spectral efficiency of SS-US falls substantially below that of subset-search methods would disprove the performance claim.

Figures

Figures reproduced from arXiv: 2604.13952 by Chathura Jayawardena, Jo\~ao Paulo S. H. Lima, Konstantinos Nikitopoulos, Marcin L. Filo.

Figure 1
Figure 1. Figure 1: System model showing the channel matrix H, when K single-antenna users are selected out of U candidates. (RBGs) represent the minimum scheduling allocation unit, each composed of one or more RBs, yielding B RBGs per slot. Let Hb ∈ CM×U denote the channel matrix in RBG b ∈ [1, B], where each column hu ∈ CM×1 represents the channel vector of user u, with ∥hu∥ denoting its norm. For each RBG, a subset of UEs … view at source ↗
Figure 2
Figure 2. Figure 2: Representing the selection of users based on their chan [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Increasing the number of parallel bases L yields higher spectral efficiency at the cost of a linear increase in complexity [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: SS-US performance for indicative α values: overly strict threshold can limit user pool and lower spectral efficiency. C. Spectral Efficiency Analysis We now assess the spectral efficiency of SS-US relative to the considered baselines, including random selection as a lower performance bound. We first analyze the impact of varying the number L of parallel random initializations for the orthonormal matrix, wi… view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of SS-US against baselines [9], [10], [12]. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
read the original abstract

The rise of Artificial Intelligence (AI)-driven services, machine-type communications, and massive Internet of Things (IoT) deployments is reshaping wireless traffic toward dense, uplink-oriented, bursty, and latency-critical patterns. In these regimes, Multi-User Multiple-Input Multiple-Output (MU-MIMO) is essential to support massive concurrent connectivity through spatial multiplexing. However, the need for frequent, low-latency scheduling decisions exposes fundamental scalability barriers in existing user selection approaches. The inherently combinatorial nature of MU-MIMO user selection leads computational complexity to grow rapidly with both the number of candidate users and spatial layers, rendering existing near-optimal heuristic methods impractical in dense and highly dynamic scenarios. This paper introduces the Space Splitting-based User Selection (SS-US) algorithm, a complexity barrier-breaking, massively parallelizable method that departs from subset-based selection by constructing orthonormal spatial bases and independently matching users to spatial directions. Simulation results across diverse MIMO configurations, channel conditions, and user densities show that SS-US reduces computational complexity by over three orders of magnitude while achieving spectral efficiency comparable to state-of-the-art practical baselines.

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.

Referee Report

1 major / 2 minor

Summary. The paper introduces the Space Splitting-based User Selection (SS-US) algorithm for MU-MIMO in massive connectivity and AI-native traffic scenarios. It departs from combinatorial subset selection by constructing orthonormal spatial bases and independently matching users to spatial directions, claiming a reduction in computational complexity by over three orders of magnitude while achieving spectral efficiency comparable to state-of-the-art practical baselines, supported by simulations across MIMO configurations, channel conditions, and user densities.

Significance. If the performance claims hold, the work would be significant for enabling scalable, low-latency user selection in dense, bursty uplink MU-MIMO systems typical of massive IoT and machine-type communications. The massively parallelizable space-splitting approach addresses a key scalability barrier in existing heuristics. The simulation validation across diverse settings provides initial evidence of practicality, though the result's impact depends on robustness to realistic channel conditions.

major comments (1)
  1. [§4] §4 (Simulation Results): The central claim of spectral efficiency comparability to baselines rests on simulations that use diverse MIMO setups and channel conditions, but do not explicitly include or report results for highly spatially correlated channels (e.g., correlation coefficients >0.5 typical in massive IoT). Without such cases or direct comparison on identical user sets, it remains unclear whether independent matching to fixed orthonormal bases incurs non-negligible residual multi-user interference, undermining the assertion that near-optimal SE is preserved alongside the complexity reduction.
minor comments (2)
  1. [Abstract] The abstract and introduction could more explicitly name the state-of-the-art practical baselines (e.g., specific greedy or semi-orthogonal algorithms) used for SE and complexity comparisons to strengthen the claims.
  2. [§3] Notation for the constructed orthonormal spatial bases and the independent matching procedure should be introduced with a clear equation or pseudocode block in the algorithm description section for improved readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive review and the opportunity to clarify aspects of our work on the SS-US algorithm. We address the major comment on simulation results for highly spatially correlated channels below, and we commit to strengthening the manuscript accordingly.

read point-by-point responses

  1. Referee: [§4] §4 (Simulation Results): The central claim of spectral efficiency comparability to baselines rests on simulations that use diverse MIMO setups and channel conditions, but do not explicitly include or report results for highly spatially correlated channels (e.g., correlation coefficients >0.5 typical in massive IoT). Without such cases or direct comparison on identical user sets, it remains unclear whether independent matching to fixed orthonormal bases incurs non-negligible residual multi-user interference, undermining the assertion that near-optimal SE is preserved alongside the complexity reduction.

    Authors: We appreciate this observation and agree that explicit results for high spatial correlation would strengthen the validation. Our original simulations span diverse channel conditions, but we did not isolate or report cases with correlation coefficients >0.5. In the revised manuscript, we will add new simulation results specifically for high-correlation scenarios (coefficients 0.6–0.9) using the same MIMO configurations and user densities as the existing figures. These will include direct SE comparisons against the baselines on identical user sets, along with an analysis of the residual interference term in the SINR expression. Theoretically, the orthonormal basis construction ensures that projections onto orthogonal directions remain independent; highly correlated users compete for the same direction and the strongest is selected, while the orthogonality limits cross-interference for the remaining assignments. We believe this will confirm that near-optimal SE is preserved. revision: yes

Circularity Check

0 steps flagged

No significant circularity; algorithmic construction validated by simulation.

full rationale

The paper defines the SS-US algorithm directly via orthonormal spatial bases and independent per-direction user matching, then reports complexity reduction and SE performance from Monte Carlo simulations across MIMO setups and channel conditions. No equations reduce the claimed gains to fitted parameters, self-referential definitions, or self-citation chains; the central claims rest on explicit algorithmic steps and external simulation benchmarks rather than by-construction equivalence to inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The contribution is a new algorithmic procedure; the abstract does not introduce or rely on additional free parameters, axioms, or invented physical entities beyond standard MIMO channel assumptions.

pith-pipeline@v0.9.0 · 5515 in / 1031 out tokens · 49061 ms · 2026-05-10T12:27:07.287785+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

  1. [1]

    Ericsson mobility report,

    Ericsson, “Ericsson mobility report,” Ericsson, 2025

    work page 2025

  2. [2]

    Global network traffic report,

    Nokia, “Global network traffic report,” Nokia, 2024

    work page 2024

  3. [3]

    Toward 6G-based metaverse: Supporting highly-dynamic deterministic multi-user extended reality services,

    H. Yu, M. Shokrnezhad, T. Taleb, R. Li, and J. Song, “Toward 6G-based metaverse: Supporting highly-dynamic deterministic multi-user extended reality services,”IEEE Network, vol. 37, no. 4, pp. 30–38, 2023

    work page 2023

  4. [4]

    Advanced manufacturing in industry 5.0: A survey of key enabling technologies and future trends,

    W. Xiang et al., “Advanced manufacturing in industry 5.0: A survey of key enabling technologies and future trends,”IEEE Trans. Ind. Inform., vol. 20, no. 2, pp. 1055–1068, 2023

    work page 2023

  5. [5]

    Vehicular networking: A survey and tutorial on requirements, architectures, challenges, standards and solutions,

    G. Karagiannis, O. Altintas, E. Ekici, G. Heijenk, B. Jarupan, K. Lin, and T. Weil, “Vehicular networking: A survey and tutorial on requirements, architectures, challenges, standards and solutions,”IEEE Commun. Sur- veys Tuts., vol. 13, no. 4, pp. 584–616, 2011

    work page 2011

  6. [6]

    State of IoT 2024: Number of connected IoT devices growing 13% to 18.8 billion globally,

    S. Sinha, “State of IoT 2024: Number of connected IoT devices growing 13% to 18.8 billion globally,” IoT Analytics, 2024. [Online]. Available: https://iot-analytics.com/number-connected-iot-devices/

    work page 2024

  7. [7]

    An overview on resource allocation techniques for multi-user MIMO systems,

    E. Casta ˜neda, A. Silva, A. Gameiro, and M. Kountouris, “An overview on resource allocation techniques for multi-user MIMO systems,”IEEE Commun. Surveys Tuts., vol. 19, no. 1, pp. 239–284, 2017

    work page 2017

  8. [8]

    Scheduling algorithms for 5G networks and beyond: Classification and survey,

    A. Mamane, M. Fattah, M. E. Ghazi, M. E. Bekkali, Y . Balboul, and S. Mazer, “Scheduling algorithms for 5G networks and beyond: Classification and survey,”IEEE Access, vol. 10, pp. 51 643–51 661, 2022

    work page 2022

  9. [9]

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

    T. Yoo and A. Goldsmith, “On the optimality of multiantenna broadcast scheduling using zero-forcing beamforming,”IEEE JSAC, vol. 24, no. 3, pp. 528–541, 2006

    work page 2006

  10. [10]

    On downlink beamforming with greedy user selection: performance analysis and a simple new algorithm,

    G. Dimic and N. Sidiropoulos, “On downlink beamforming with greedy user selection: performance analysis and a simple new algorithm,”IEEE Trans. Signal Process., vol. 53, no. 10, pp. 3857–3868, 2005

    work page 2005

  11. [11]

    Contiguous multi-user scheduling and power control for 5G-NR uplink,

    A. Sengupta, B. Mondal, V . Sergeev, A. Davydov, and A. Papathanassiou, “Contiguous multi-user scheduling and power control for 5G-NR uplink,” inProc. 55th Annu. Conf. Inf. Sci. Syst. (CISS), 2021, pp. 1–6

    work page 2021

  12. [12]

    mCore+: A real-time design achieving∼500µs scheduling for 5G MU-MIMO systems,

    Y . Chen, Y . Wu, Y . T. Hou, and W. Lou, “mCore+: A real-time design achieving∼500µs scheduling for 5G MU-MIMO systems,”IEEE Trans. Mobile Comput., vol. 22, no. 12, pp. 7249–7265, 2023

    work page 2023

  13. [13]

    Study on channel model for frequencies from 0.5 to 100 GHz,

    3GPP, “Study on channel model for frequencies from 0.5 to 100 GHz,” 3rd Generation Partnership Project, TS 38.901 V19.1.0, Sep. 2025

    work page 2025

  14. [14]

    NR; Physical layer procedures for data,

    ——, “NR; Physical layer procedures for data,” 3rd Generation Partner- ship Project, TS 38.214 V19.1.0, Sep. 2025

    work page 2025