Pith. sign in

REVIEW 3 major objections 6 minor 57 references

Graph Neural Network Aided Detection for the Multi-User Multi-Dimensional Index Modulated Uplink

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Graph neural networks fused with message-passing detectors (GNN-AMP and GEPNet) approach the maximum-likelihood performance on a compressed sensing-aided space-frequency index modulation uplink, while a single GNN training suffices for…

desk verdict Solid incremental application of GNN-aided MP detectors to a new CS-SFIM variant; the 'train once for any user count' claim is unsupported as written. read the letter →

arxiv 2505.21343 v1 pith:IJCBZ63T submitted 2025-05-27 eess.SP

Xinyu Feng , Mohammed EL-Hajjar , Chao Xu , Lajos Hanzo This is my paper · ORCID
classification eess.SP
keywords IndexModulationMulti-UserMIMOGraphNeuralNetworkMessagePassingApproximateExpectationPropagationCompressedSensingSpace-Frequency
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper claims that adding a graph neural network module to two message-passing detectors, AMP and EP, yields near-optimal detection for a compressed sensing-aided space-frequency index modulation (CS-SFIM) uplink with many users. The proposed GNN-AMP and GEPNet detectors are shown to outperform conventional MP and GNN detectors and to come within a few tenths of a dB of the maximum-likelihood (ML) bound, while avoiding ML's exponential complexity. The paper further claims that a single GNN training suffices across different numbers of active users, because the underlying detector is permutation-equivariant. If true, this would give large-scale multi-user MIMO a low-complexity, near-optimal receiver that does not need retraining as user load changes.

What carries the argument

The central object is the joint factor graph of the space-frequency domain, where variable nodes are the elements of the equivalent sparse transmit vector x and factor nodes are the received observations across receive antennas and subcarriers. The load-bearing mechanism is the GNN module that replaces the Gaussian approximation of the cavity distribution in AMP and EP: node attributes are the MP-computed mean and variance, edge attributes are the channel cross-correlations and noise variance, and an MLP-plus-GRU message-passing scheme iteratively refines the posterior. A claimed permutation equivariance of this message-passing graph is what licenses the single-training/many-users property.

What would settle it

Measure BER of GNN-AMP and GEPNet with imperfect channel estimates (e.g., 10% normalized MSE) or with a number of active users outside the training range, and compare against ML; a gap that grows by more than a few dB, or a broken equivariance under user permutation, would refute the near-ML and train-once claims.

Watch

Extended reading notes

Core claim

The paper designs a CS-SFIM transceiver in which each user maps bits to subcarrier activation indices and antenna activation indices separately, forming a sparse space-frequency matrix per subcarrier group. Treating the entire multi-user SF matrix as a joint factor graph, the authors build AMP and EP detectors, then interleave a graph neural network into each iteration: the GNN takes the cavity parameters (mean, variance) of the MP algorithm as node attributes and refines the posterior distribution that the MP module uses. The resulting GEPNet detector reaches a BER of $10^{-4}$ at an SNR within about 0.3-0.35 dB of ML in the 4-user configurations, and both GNN-AMP and GEPNet remain the best-performing practical detectors in the 16-user configurations. The train-once claim is supported by the permutation equivariance of the GNN message-passing architecture, which lets the same trained weights act on any subset of users.

Load-bearing premise

The results assume perfect channel knowledge and perfect timing at the base station, and the claim that one training works for any number of users rests on a symmetry property that the paper states but does not fully prove.

Editorial extensions

If this is right

  • GEPNet achieves BER within about 0.3 dB of the maximum-likelihood detector in the 4-user configurations, while GNN-AMP remains within about 0.3-0.35 dB with lower complexity.
  • Both GNN-aided detectors outperform the conventional EP, AMP, GNN-MMSE, and DNN detectors by several dB in all four tested configurations, with the gap widening as user count grows.
  • A single GNN training, performed on a mixture of active-user counts, suffices for any number of active users from 1 to the maximum supported, without retraining per user count.
  • Increasing the number of receive antennas improves all GNN-aided detectors, and the near-ML performance is maintained in the large-scale 16-user, 64-128 receive-antenna uplink.
  • The detector family tolerates user activity variation, degrading gracefully as the number of active users grows.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The permutation-equivariance argument, made rigorous, would imply the learned message-passing weights depend on channel statistics rather than on user identity; a direct stress test is to apply the detector to a user count outside the training range.
  • Treating channel estimation error as part of the edge attributes and training on noisy channel realizations could extend the near-ML gains to imperfect CSI; the paper does not explore this.
  • Because the GNN module refines posterior distributions rather than hard decisions, it could be coupled with channel coding for joint detection and decoding, as the conclusion itself suggests.
  • The complexity numbers in the paper count multiplications only; a fairer systems comparison would include memory, latency, and training cost for the GNN modules.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The manuscript proposes CS-SFIM for the large-scale multiuser MIMO uplink, in which each user's information bits separately select active transmit antennas and active subcarriers and are also mapped to APM symbols. The BS detection problem is represented as a factor graph, and the authors develop AMP and EP detectors and then two GNN-aided variants, GNN-AMP and GEPNet, together with a GNN-MMSE baseline. The paper claims that the GNN-aided detectors approach ML performance at reduced complexity and that a single GNN training suffices for systems with a variable number of users, verified by an alleged permutation-equivariance proof. The evidence is BER simulations over Rayleigh channels for U=4 and U=16 users, plus complexity comparisons and a robustness experiment across active user counts.

Significance. If the performance claims hold, the paper makes a useful contribution to index-modulation-based massive MIMO uplink detection: it extends compressed-sensing-aided multidimensional IM to a multiuser setting, formulates the factor graph for the resulting sparse signal, and demonstrates that unrolling AMP/EP with GNN modules can give large BER gains over the classical counterparts. The ML benchmark is an independent upper bound rather than a training target for the GNN, so the overall comparison is not circular. The complexity table and the evaluation across different numbers of active users are informative. However, the headline 'train once for any number of users' claim is supported neither by the promised proof nor by the presented experiment, and the EP/GEPNet derivation contains a mean/variance labeling inconsistency; these issues must be resolved before the central claims can be accepted.

major comments (3)
  1. [Section III-A2, Eqs. (46)-(47)] Equations (46) and (47) assign the two parameters of the cavity distribution in (45) in an internally inconsistent way. For the quotient N(mu_i,Sigma_ii)/N(eta_i/V_i,1/V_i), the standard Gaussian identity gives variance v_o,i = Sigma_ii/(1 - Sigma_ii*V_i) and mean m_o,i = v_o,i*(mu_i/Sigma_ii - eta_i). The text instead labels the variance expression as m_o,i in (46) and the mean expression as v_o,i in (47). These labels feed Algorithm 2 (line 4), Algorithm 4 (line 4), and the GEPNet node attribute in (69), so an implementation following the text would use the wrong cavity statistics. The authors should either swap the labels/expressions or explicitly define a nonstandard convention and use it consistently; as written, the GEPNet derivation is inconsistent and the near-ML GEPNet results cannot be checked against the equations.
  2. [Section I, contribution 4; Section IV, Fig. 11] Contribution 4 in Section I claims that a single GNN training works for any number of users and states that this is verified by a mathematical proof of permutation equivariance, but no such proof appears in Sections II-IV. The only supporting experiment, Fig. 11, explicitly trains all GNN-aided methods on a mixture of one active user up to the maximum number of users and then evaluates the same range of Ka; this is interpolation over the training distribution, not a test of generalization to an unseen number of users. In addition, permutation equivariance of the GNN message-passing module concerns relabeling nodes inside a fixed graph, whereas changing Umax changes the graph size, the dimensions of Hg, and the EP/AMP linear modules in (42)-(43) and Algorithm 1. To support the train-once claim, the authors should either provide a proof that covers insertion/removal of users, not merely node permutations, or test on unseen Umax/Ka settings, for example training at Umax=4 and testing at Umax=8 and 16.
  3. [Section IV] Section IV reports BER curves for the GNN and DNN detectors but does not state the training setup: number of training samples, SNR values or distribution used for training, optimizer, learning rate, number of epochs, train/validation split, or the DNN architecture and its label generation beyond 'trained by the data of ML detection'. Because the central claims are established purely by simulation, these details are needed to assess sensitivity and reproducibility. Please add a training-configuration table and, if possible, a statement on code/data availability.
minor comments (6)
  1. [Section II-A3, Eq. (2)] The rate expression contains a stray ']' and unbalanced parentheses ('K log2 C(Nt, Na) +] + NaK log2 L'); it should be rewritten.
  2. [Section III-A1, Eq. (15)] The ML objective uses uppercase Yg while the received signal is defined as yg in Eq. (7); use consistent notation.
  3. [Section II-A, Eqs. (3)-(5)] The model introduces a per-user noise w_i^u and then defines w_i = sum_u w_i^u in (5), which inflates the noise variance by U unless the individual variances are scaled; this is inconsistent with the conventional per-antenna noise model later used in (6). Please clarify the noise model.
  4. [Section IV, Scheme 5] The scheme description lists 64 RAs, but the accompanying text says 'employ a 16x16 MIMO size' and Fig. 12 is described in terms of U=4; please resolve this inconsistency.
  5. [Notation throughout] The paper uses 'AMP-GNN' and 'GNN-AMP' interchangeably (e.g., contribution 3, Section III-C3, Fig. 7, and Table III); pick one consistent name.
  6. [Section III-D and Table III] The claimed DNN complexity O(Nh1 Nh2) is missing the dependence on the input/output dimensions and the number of layers; specify the complexity in terms of the architecture actually used.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the GNN detectors are evaluated against an independent ML benchmark, and the train-once claim, while under-supported, is not circular.

full rationale

The paper's central performance comparisons (GEPNet and GNN-AMP versus EP, AMP, and ML) do not reduce to their own inputs. ML detection is an independent exhaustive-search benchmark defined in Eq. (15), and it is not a function of the trained GNN parameters. The GNN detectors are model-driven augmentations: GNN-AMP replaces the Gaussian approximation in AMP with a learned distribution (Section III-C3, Eqs. 66-68), and GEPNet replaces the EP cavity distribution with a GNN readout (Section III-C4, Eqs. 69-72). Neither architecture's output is defined to equal the ML result, and no fitted parameter is renamed as a prediction. The CS-SFIM system model is taken from prior work [27], but that citation supplies the channel and modulation model rather than the detectors' performance. The only notable caveat is the 'train once for any number of users' claim in Section I, contribution 4: the promised permutation-equivariance proof is not provided, and Fig. 11's training protocol ('All GNN-aided methods are trained by a scheme that applies a mixture of a single active user to the maximum number of users to see the full range') evaluates interpolation over the trained user-count range rather than generalization to an unseen user count. This is a missing proof and an in-sample evaluation, which is a correctness and generalization risk, but it is not a circular reduction of the central claim.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

Central claims rest on a system model with perfect CSI and synchronization, Gaussian message approximations, and an unproved permutation equivariance property of the GNN. The only explicitly fitted numbers are GNN weights and a few damping coefficients; their values are not reported, which makes the simulation results hard to audit.

free parameters (4)
  • GNN weights and biases (W1, b1, W2, b2, D, U, R) = not reported
    Trained on simulated data; exact values and training setup are omitted, so the central near-ML results cannot be reproduced from the paper alone.
  • GNN architecture hyperparameters (L, T, N_h1, N_h2, N_u) = not reported
    Number of GNN layers, message-passing iterations, and hidden sizes are never specified in Section III-C or Section IV.
  • EP damping coefficient epsilon = 0.05 (Algorithm 2), 0.1 (Algorithm 4)
    Chosen by hand to stabilize updates; no search or sensitivity analysis is reported.
  • Training data distribution over active user counts = mixture of single active user to maximum users
    Section IV states the training mixture for Fig. 11, but the exact proportions, SNR values, and sample counts are not given, and this distribution directly determines the claimed user-count robustness.
assumptions (5)
  • domain assumption Perfect CSI and perfect synchronization at the base station
    Footnotes 3 and 4 in Section II state the CSI and synchronization assumptions; all BER curves in Section IV use this model. Imperfect CSI is known to degrade AMP and GNN detectors (Section I).
  • domain assumption User separability is achieved by unique space-frequency signatures imposed by each user's channel
    Section III opening; needed for the stacked model y = Hs + w to be invertible by the detectors.
  • standard math Gaussian approximation of messages and cavity distributions in AMP and EP
    Section III-B, Eqs. (29)-(30) and (39)-(45); a standard approximation, but its accuracy controls detector performance in high-MUI scenarios.
  • domain assumption GNN aggregation is permutation equivariant over variable numbers of users
    Contribution 4 states a mathematical proof, but no proof is given; the claim that one trained model works for any user count depends on this property.
  • ad hoc to paper GNN hyperparameters and damping coefficients are adequate across all schemes
    Algorithm 2 and Algorithm 4 fix epsilon by hand, and GNN sizes are not reported; no sensitivity analysis is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Graph Neural Network Aided Detection for the Multi-User Multi-Dimensional Index Modulated Uplink." pith.science (2026). https://pith.science/paper/IJCBZ63T

@misc{pith2026250521343,
  author       = {Pith},
  title        = {Pith review of: Graph Neural Network Aided Detection for the Multi-User Multi-Dimensional Index Modulated Uplink},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IJCBZ63T}},
  note         = {Machine review of arXiv:2505.21343}
}
read the original abstract

The concept of Compressed Sensing-aided Space-Frequency Index Modulation (CS-SFIM) is conceived for the Large-Scale Multi-User Multiple-Input Multiple-Output Uplink (LS-MU-MIMO-UL) of Next-Generation (NG) networks. Explicitly, in CS-SFIM, the information bits are mapped to both spatial- and frequency-domain indices, where we treat the activation patterns of the transmit antennas and of the subcarriers separately. Serving a large number of users in an MU-MIMO-UL system leads to substantial Multi-User Interference (MUI). Hence, we design the Space-Frequency (SF) domain matrix as a joint factor graph, where the Approximate Message Passing (AMP) and Expectation Propagation (EP) based MU detectors can be utilized. In the LS-MU-MIMO-UL scenario considered, the proposed system uses optimal Maximum Likelihood (ML) and Minimum Mean Square Error (MMSE) detectors as benchmarks for comparison with the proposed MP-based detectors. These MP-based detectors significantly reduce the detection complexity compared to ML detection, making the design eminently suitable for LS-MU scenarios. To further reduce the detection complexity and improve the detection performance, we propose a pair of Graph Neural Network (GNN) based detectors, which rely on the orthogonal AMP (OAMP) and on the EP algorithm, which we refer to as the GNN-AMP and GEPNet detectors, respectively. The GEPNet detector maximizes the detection performance, while the GNN-AMP detector strikes a performance versus complexity trade-off. The GNN is trained for a single system configuration and yet it can be used for any number of users in the system. The simulation results show that the GNN-based detector approaches the ML performance in various configurations.

Figures

Figures reproduced from arXiv: 2505.21343 by the authors.

Figure 1
Figure 1. MU CS-SFIM UL transceiver architecture, where the BS has [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. structure and Example of CS-SFIM modulation at [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Factor graph for CS-SFIM xˆi = arg max X x∈SΩ p(x|y). (20) According to Bayes’ theorem, we can have p(x|y) = p(y|x)p(x) p(y) ∝ p(y|x)p(x), (21) where p(x) = QΩ i=1 p(xi) is the joint a priori probability function of all the symbols in the SF matrix domain and p(y) = P x∈SΩ p(y|x) is the PDF of the received signal y represented as: p(y|x) = Y Φ j=1 p(yj |x), (22) where Φ = NrNf is the size of the received signal vect… view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: Aggregation process of the GNN a(t) 1 a(t) k a(t) Ω u(L u k (L 1 uL Ω p (t) GNN (a1) p(t) GNN (aΩ) R(u (L) k ) [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Readout process of the GNN The message m (l) i is used for computing the nodes’ hidden vector u (l) i as follow: g (l) i = U(g (l−1) i , m (l) i ), (60) u (l) i = W2g (l) i + b2, (61) where the function U is constituted by a Gated Recurrent Unit (GRU) [60] based networ…
Figure 7
Figure 7. Figure 7: The structure of the AMP-GNN network. 2) GNN-MMSE: In this section, to improve the prior infor￾mation, we incorporate the MMSE a posteriori as the a priori p(x) derived from (16) such that p(xi) = 1 πhcii exp [− (zi − xi) 2 hcii ], (65) where zi is the i-th element of …
Figure 8
Figure 8. Figure 8: The structure of the GEPnet. Algorithm 4: GEPnet Detector Input: y, H¯ , σ2 , Ω, Φ, transmit power Es Initialization: ˆx t=1 = 0, η t=1 = 0, V t=1 = 1 Es I,ϵ = 0.1, g (0) = 0 Iteration: 1 for t = 1, 2, · · · , T do 2 EP Observation Module 3 Compute Σ and µ based on (42…
Figure 9
Figure 9. Figure 9: BER performance comparison of the different MU CS [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: BER performance comparison of the different MU [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: BER performance versus the number of active users [PITH_FULL_IMAGE:figures/full_fig_p015_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 54 canonical work pages

  1. [27]

    Two-Dimensional Index Modulation for the Large-Scale Multi-User MIMO Uplink,

    S. Lu, M. El-Hajjar, and L. Hanzo, “Two-Dimensional Index Modulation for the Large-Scale Multi-User MIMO Uplink,” IEEE Transactions on Vehicular Technology, vol. 68, no. 8, pp. 7904–7918, 2019

    work page 2019

  2. [2]

    Fifty Years of MIMO Detection: The Road to Large-Scale MIMOs,

    S. Yang and L. Hanzo, “Fifty Years of MIMO Detection: The Road to Large-Scale MIMOs,” IEEE Communications Surveys and Tutorials , vol. 17, no. 4, pp. 1941–1988, 2015

    work page 1941

  3. [3]

    Massive MIMO Detection Techniques: A Survey,

    M. A. Albreem, M. Juntti, and S. Shahabuddin, “Massive MIMO Detection Techniques: A Survey,” IEEE Communications Surveys and Tutorials, vol. 21, no. 4, pp. 3109–3132, 2019

    work page 2019

  4. [4]

    Survey on Massive MIMO: Technol- ogy, Challenges, Opportunities and Benefits,

    K. K. Vaigandla and D. N. Venu, “Survey on Massive MIMO: Technol- ogy, Challenges, Opportunities and Benefits,” 2021

    work page 2021

  5. [5]

    Space Modulation On Wireless Fading Chan- nels,

    Y . Chau and S.-H. Yu, “Space Modulation On Wireless Fading Chan- nels,” in IEEE 54th Vehicular Technology Conference. VTC Fall 2001. Proceedings (Cat. No.01CH37211) , vol. 3, 2001, pp. 1668–1671 vol.3

    work page 2001

  6. [6]

    50 Years of Permutation, Spatial and Index Modulation: From Classic RF to Visible Light Communications and Data Storage,

    N. Ishikawa, S. Sugiura, and L. Hanzo, “50 Years of Permutation, Spatial and Index Modulation: From Classic RF to Visible Light Communications and Data Storage,” IEEE Communications Surveys and Tutorials, vol. 20, no. 3, pp. 1905–1938, 2018

    work page 1905

  7. [7]

    Spatial Modulation,

    R. Y . Mesleh, H. Haas, S. Sinanovic, C. W. Ahn, and S. Yun, “Spatial Modulation,” IEEE Transactions on Vehicular Technology, vol. 57, no. 4, pp. 2228–2241, 2008

    work page 2008

  8. [8]

    Spatial Modulation and Space-Time Shift Keying: Optimal Performance at a Reduced Detection Complexity,

    C. Xu, S. Sugiura, S. X. Ng, and L. Hanzo, “Spatial Modulation and Space-Time Shift Keying: Optimal Performance at a Reduced Detection Complexity,” IEEE Transactions on Communications, vol. 61, no. 1, pp. 206–216, 2013

    work page 2013

Show all 57 references
  1. [9]

    On Massive Spatial Modulation MIMO: Spectral Efficiency Analysis and Optimal System Design,

    L. He, J. Wang, and J. Song, “On Massive Spatial Modulation MIMO: Spectral Efficiency Analysis and Optimal System Design,” in 2016 IEEE Global Communications Conference (GLOBECOM) , 2016, pp. 1–6

    2016

  2. [10]

    Design Guidelines for Spatial Modulation,

    P. Yang, M. Di Renzo, Y . Xiao, S. Li, and L. Hanzo, “Design Guidelines for Spatial Modulation,” IEEE Communications Surveys and Tutorials , vol. 17, no. 1, pp. 6–26, 2015

    2015

  3. [11]

    On Generalized Spatial Modulation,

    T. Datta and A. Chockalingam, “On Generalized Spatial Modulation,” in 2013 IEEE Wireless Communications and Networking Conference (WCNC), 2013, pp. 2716–2721

    2013

  4. [12]

    Gener- alized Spatial Modulation in Large-Scale Multiuser MIMO Systems,

    T. Lakshmi Narasimhan, P. Raviteja, and A. Chockalingam, “Gener- alized Spatial Modulation in Large-Scale Multiuser MIMO Systems,” IEEE Transactions on Wireless Communications , vol. 14, no. 7, pp. 3764–3779, 2015

    2015

  5. [13]

    Index Modulation Techniques For 5G Wireless Networks,

    E. Basar, “Index Modulation Techniques For 5G Wireless Networks,” IEEE Communications Magazine , vol. 54, no. 7, pp. 168–175, 7 2016

    2016

  6. [14]

    Subcarrier-Index Modulation OFDM,

    R. Abu-alhiga and H. Haas, “Subcarrier-Index Modulation OFDM,” 2009 IEEE 20th International Symposium on Personal, Indoor and Mobile Radio Communications , pp. 177–181, 2009

    2009

  7. [15]

    On the Achievable Rate of OFDM With Index Modulation,

    M. Wen, X. Cheng, M. Ma, B. Jiao, and H. V . Poor, “On the Achievable Rate of OFDM With Index Modulation,” IEEE Transactions on Signal Processing, vol. 64, no. 8, pp. 1919–1932, 2016

    1919

  8. [16]

    Generalization of Orthogonal Frequency Division Multiplexing With Index Modulation,

    R. Fan, Y . J. Yu, and Y . L. Guan, “Generalization of Orthogonal Frequency Division Multiplexing With Index Modulation,” IEEE Trans- actions on Wireless Communications , vol. 14, no. 10, pp. 5350–5359, 2015

    2015

  9. [17]

    Uplink and Downlink Transceiver Design for OFDM with Index Modulation in Multi-user Networks,

    M. Y ¨uzgec ¸cio˘glu and E. Jorswieck, “Uplink and Downlink Transceiver Design for OFDM with Index Modulation in Multi-user Networks,” in 2017 IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC) , 2017, pp. 1–5

    2017

  10. [19]

    Compressed Sensing,

    D. L. Donoho, “Compressed Sensing,” IEEE Transactions on Informa- tion Theory, vol. 52, no. 4, pp. 1289–1306, apr 2006

    2006

  11. [20]

    Gener- alized Space-and-Frequency Index Modulation,

    T. Datta, H. S. Eshwaraiah, A. Chockalingam, and S. Member, “Gener- alized Space-and-Frequency Index Modulation,” IEEE Transactions on Vehicular Technology, vol. 65, no. 7, pp. 4911–4924, 2016

    2016

  12. [21]

    Hierarchical Multi- Functional Layered Spatial Modulation,

    I. A. Hemadeh, M. El-Hajjar, and L. Hanzo, “Hierarchical Multi- Functional Layered Spatial Modulation,” IEEE Access, vol. 6, pp. 9492– 9533, 2018

    2018

  13. [22]

    Multidimen- sional Index Modulation in Wireless Communications,

    B. Shamasundar, S. Bhat, S. Jacob, and A. Chockalingam, “Multidimen- sional Index Modulation in Wireless Communications,” IEEE Access , vol. 6, pp. 589–604, 2018

    2018

  14. [23]

    Mul- tidomain Index Modulation for Vehicular and Railway Communications: A Survey of Novel Techniques,

    P. Yang, Y . Xiao, Y . L. Guan, M. Di Renzo, S. Li, and L. Hanzo, “Mul- tidomain Index Modulation for Vehicular and Railway Communications: A Survey of Novel Techniques,” IEEE Vehicular Technology Magazine, vol. 13, no. 3, pp. 124–134, 2018

    2018

  15. [24]

    Space-, Time- and Frequency- Domain Index Modulation for Next-Generation Wireless: A Unified Single-/Multi-Carrier and Single-/Multi-RF MIMO Framework,

    C. Xu, Y . Xiong, N. Ishikawa, R. Rajashekar, S. Sugiura, Z. Wang, S.-X. Ng, L.-L. Yang, and L. Hanzo, “Space-, Time- and Frequency- Domain Index Modulation for Next-Generation Wireless: A Unified Single-/Multi-Carrier and Single-/Multi-RF MIMO Framework,” IEEE Transactions on...

    2021

  16. [25]

    Compressed Sensing-Aided Multi-Dimensional Index Modulation,

    S. Lu, I. A. Hemadeh, M. El-Hajjar, and L. Hanzo, “Compressed Sensing-Aided Multi-Dimensional Index Modulation,” IEEE Transac- tions on Communications , vol. 67, no. 6, pp. 4074–4087, 2019

    2019

  17. [26]

    Near-Instantaneously Adaptive Learning-Assisted and Compressed Sensing-Aided Joint Multi- Dimensional Index Modulation,

    X. Feng, M. El-Hajjar, C. Xu, and L. Hanzo, “Near-Instantaneously Adaptive Learning-Assisted and Compressed Sensing-Aided Joint Multi- Dimensional Index Modulation,” IEEE Open Journal of Vehicular Tech- nology, vol. 4, pp. 893–912, 2023

    2023

  18. [28]

    Message Passing Algo- rithms for Compressed Sensing,

    D. L. Donoho, A. Maleki, and A. Montanari, “Message Passing Algo- rithms for Compressed Sensing,” Proceedings of the National Academy of Sciences, vol. 106, no. 45, pp. 18 914–18 919, 2009

    2009

  19. [29]

    Optimality of Large MIMO Detection via Approximate Message Passing,

    C. Jeon, R. Ghods, A. Maleki, and C. Studer, “Optimality of Large MIMO Detection via Approximate Message Passing,” in 2015 IEEE International Symposium on Information Theory (ISIT), 2015, pp. 1227– 1231

    2015

  20. [32]

    Generalized Space-Frequency Index Modulation: Low-Complexity Encoding and Detection,

    B. Chakrapani, T. L. Narasimhan, and A. Chockalingam, “Generalized Space-Frequency Index Modulation: Low-Complexity Encoding and Detection,” in 2015 IEEE Globecom Workshops (GC Wkshps) , 2015, pp. 1–6

    2015

  21. [33]

    Orthogonal AMP,

    J. Ma and L. Ping, “Orthogonal AMP,” IEEE Access, vol. 5, pp. 2020– 2033, 2017

    2020

  22. [34]

    Expectation Propagation Detection for High-Order High-Dimensional MIMO Systems,

    J. C ´espedes, P. M. Olmos, M. S ´anchez-Fern´andez, and F. Perez-Cruz, “Expectation Propagation Detection for High-Order High-Dimensional MIMO Systems,” IEEE Transactions on Communications, vol. 62, no. 8, pp. 2840–2849, 2014

    2014

  23. [35]

    Expecta- tion Propagation Detector for Extra-Large Scale Massive MIMO,

    H. Wang, A. Kosasih, C.-K. Wen, S. Jin, and W. Hardjawana, “Expecta- tion Propagation Detector for Extra-Large Scale Massive MIMO,” IEEE Transactions on Wireless Communications , vol. 19, no. 3, pp. 2036– 2051, 2020

    2020

  24. [36]

    Model-Driven Deep Learning for MIMO Detection,

    H. He, C.-K. Wen, S. Jin, and G. Y . Li, “Model-Driven Deep Learning for MIMO Detection,” IEEE Transactions on Signal Processing, vol. 68, pp. 1702–1715, 2020

    2020

  25. [37]

    Model-Driven Deep Learning for Physical Layer Communications,

    H. He, S. Jin, C.-K. Wen, F. Gao, G. Y . Li, and Z. Xu, “Model-Driven Deep Learning for Physical Layer Communications,” IEEE Wireless Communications, vol. 26, no. 5, pp. 77–83, 2019

    2019

  26. [38]

    Learning to Detect,

    N. Samuel and T. Diskin, “Learning to Detect,” IEEE Transactions on Signal Processing, vol. 67, no. 10, pp. 2554–2564, may 2019

    2019

  27. [39]

    Graph Neural Networks for Massive MIMO Detection,

    A. Scotti, N. N. Moghadam, D. Liu, K. Gafvert, and J. Huang, “Graph Neural Networks for Massive MIMO Detection,” arXiv preprint arXiv:2007.05703, 2020. 17

    2007 arXiv

  28. [40]

    Graph Neural Networks for Wireless Communications: From Theory to Practice,

    Y . Shen, J. Zhang, S. H. Song, and K. B. Letaief, “Graph Neural Networks for Wireless Communications: From Theory to Practice,” IEEE Transactions on Wireless Communications , vol. 22, no. 5, pp. 3554–3569, 2023

    2023

  29. [41]

    GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection,

    H. He, A. Kosasih, X. Yu, J. Zhang, S. Song, W. Hardjawana, and K. B. Letaief, “GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection,” in 2023 IEEE Wireless Com- munications and Networking Conference (WCNC) , 2023, pp. 1–6

    2023

  30. [42]

    Graph Neural Network Aided Expectation Propagation Detector for MU-MIMO Systems,

    A. Kosasih, V . Onasis, W. Hardjawana, V . Miloslavskaya, V . Andrean, J.-S. Leu, and B. Vucetic, “Graph Neural Network Aided Expectation Propagation Detector for MU-MIMO Systems,” in 2022 IEEE Wireless Communications and Networking Conference (WCNC), 2022, pp. 1212– 1217

    2022

  31. [43]

    Graph Neural Network Aided MU-MIMO Detectors,

    A. Kosasih, V . Onasis, V . Miloslavskaya, W. Hardjawana, V . Andrean, and B. Vucetic, “Graph Neural Network Aided MU-MIMO Detectors,” IEEE Journal on Selected Areas in Communications , vol. 40, no. 9, pp. 2540–2555, 2022

    2022

  32. [44]

    Low- Complexity Detection for Index Modulation Multiple Access,

    J. Li, Q. Li, S. Dang, M. Wen, X.-Q. Jiang, and Y . Peng, “Low- Complexity Detection for Index Modulation Multiple Access,” IEEE Wireless Communications Letters, vol. 9, no. 7, pp. 943–947, 2020

    2020

  33. [45]

    Multi-User Spatial Modulation MIMO,

    S. Narayanan, M. J. Chaudhry, A. Stavridis, M. Di Renzo, F. Graziosi, and H. Haas, “Multi-User Spatial Modulation MIMO,” in 2014 IEEE Wireless Communications and Networking Conference (WCNC) , 2014, pp. 671–676

    2014

  34. [46]

    Non-Coherent Multi-User Detection Based on Expectation Propagation,

    K.-H. Ngo, M. Guillaud, A. Decurninge, S. Yang, S. Sarkar, and P. Schniter, “Non-Coherent Multi-User Detection Based on Expectation Propagation,” in 2019 53rd Asilomar Conference on Signals, Systems, and Computers, 2019, pp. 2092–2096

    2019

  35. [47]

    Spatial Modulation Orthogonal Frequency Division Multiplexing with Subcarrier Index Modulation for V2X Communications,

    Y . Liu, M. Zhang, H. Wang, and X. Cheng, “Spatial Modulation Orthogonal Frequency Division Multiplexing with Subcarrier Index Modulation for V2X Communications,” in 2016 International Confer- ence on Computing, Networking and Communications (ICNC) , 2016, pp. 1–5

    2016

  36. [48]

    Compressed Sensing Improves the Performance of Subcarrier Index-Modulation-Assisted OFDM,

    H. Zhang, L.-L. Yang, and L. Hanzo, “Compressed Sensing Improves the Performance of Subcarrier Index-Modulation-Assisted OFDM,” IEEE Access, vol. 4, pp. 7859–7873, 2016

    2016

  37. [49]

    Multiuser MIMO-OFDM for Next-Generation Wireless Systems,

    M. Jiang and L. Hanzo, “Multiuser MIMO-OFDM for Next-Generation Wireless Systems,” Proceedings of the IEEE , vol. 95, no. 7, pp. 1430– 1469, 2007

    2007

  38. [50]

    Multiuser Detection for Nonlinear MIMO Uplink,

    S. Chen, S. X. Ng, E. F. Khalaf, A. Morfeq, and N. D. Alotaibi, “Multiuser Detection for Nonlinear MIMO Uplink,” IEEE Transactions on Communications, vol. 68, no. 1, pp. 207–219, 2020

    2020

  39. [51]

    Iterative Synchronization of Mul- tiuser Ultra-Wideband Signals,

    E. Ekrem, M. Koca, and H. Delic ¸, “Iterative Synchronization of Mul- tiuser Ultra-Wideband Signals,” IEEE Transactions on Wireless Com- munications, vol. 9, no. 10, pp. 3040–3051, 2010

    2010

  40. [52]

    Joint transmitter-receiver optimization in synchronous multiuser communications over multipath channels,

    W. M. Jang, B. V ojcic, and R. Pickholtz, “Joint transmitter-receiver optimization in synchronous multiuser communications over multipath channels,” IEEE Transactions on Communications , vol. 46, no. 2, pp. 269–278, 1998

    1998

  41. [53]

    Factor Graphs and the Sum-Product Algorithm,

    F. Kschischang, B. Frey, and H.-A. Loeliger, “Factor Graphs and the Sum-Product Algorithm,” IEEE Transactions on Information Theory , vol. 47, no. 2, pp. 498–519, 2001

    2001

  42. [54]

    Approximate Message Passing Algorithms for Low Complexity OFDM-IM Detec- tion,

    Z. Sui, S. Yan, H. Zhang, L.-L. Yang, and L. Hanzo, “Approximate Message Passing Algorithms for Low Complexity OFDM-IM Detec- tion,” IEEE Transactions on Vehicular Technology , vol. 70, no. 9, pp. 9607–9612, 2021

    2021

  43. [55]

    Approximate Message Passing Detector for Index Modulation With Multiple Active Resources,

    L. Wei, J. Zheng, and Q. Liu, “Approximate Message Passing Detector for Index Modulation With Multiple Active Resources,” IEEE Transac- tions on Vehicular Technology, vol. 68, no. 1, pp. 972–976, 2019

    2019

  44. [56]

    Expectation-Maximization Gaussian-Mixture Approximate Message Passing,

    J. P. Vila and P. Schniter, “Expectation-Maximization Gaussian-Mixture Approximate Message Passing,” IEEE Transactions on Signal Process- ing, vol. 61, no. 19, pp. 4658–4672, 2013

    2013

  45. [57]

    Generalized Approximate Message Passing for Estimation with Random Linear Mixing,

    S. Rangan, “Generalized Approximate Message Passing for Estimation with Random Linear Mixing,” in 2011 IEEE International Symposium on Information Theory Proceedings , 2011, pp. 2168–2172

    2011

  46. [58]

    Expectation Propagation for Approximate Bayesian Infer- ence,

    T. P. Minka, “Expectation Propagation for Approximate Bayesian Infer- ence,” arXiv preprint arXiv:1301.2294 , 2013

    2013 arXiv

  47. [59]

    A review of graph neural networks: concepts, architectures, techniques, challenges, datasets, applications, and future directions,

    B. Khemani, S. Patil, K. Kotecha, and S. Tanwar, “A review of graph neural networks: concepts, architectures, techniques, challenges, datasets, applications, and future directions,” Journal of Big Data , vol. 11, no. 1, p. 18, 2024

    2024

  48. [60]

    Learning Phrase Representations us- ing RNN Encoder-Decoder for Statistical Machine Translation,

    K. Cho, B. Van Merri ¨enboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y . Bengio, “Learning Phrase Representations us- ing RNN Encoder-Decoder for Statistical Machine Translation,” arXiv preprint arXiv:1406.1078, 2014

    2014 arXiv

  49. [61]

    Near-instantaneously adaptive learning-assisted and compressed sensing-aided joint multi- dimensional index modulation,

    X. Feng, M. El-Hajjar, C. Xu, and L. Hanzo, “Near-instantaneously adaptive learning-assisted and compressed sensing-aided joint multi- dimensional index modulation,” IEEE Open Journal of Vehicular Tech- nology, vol. 4, pp. 893–912, 2023

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.