Neural CRC Prediction for 5G NR URLLC

Prashanth Murthy

REVIEW 3 major objections 5 minor 13 references

Neural CRC Prediction for 5G NR URLLC

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A neural network can predict whether a 5G uplink block will pass its CRC check before decoding finishes, cutting retransmission delay from 5 ms to 0.5 ms.

desk verdict Early CRC prediction for URLLC is a genuinely new and worthwhile idea, but the latency claim is unsupported by the full-bandwidth measurements. read the letter →

arxiv 2608.06230 v1 pith:SY4RCRRV submitted 2026-08-06 eess.SP

Prashanth Murthy This is my paper · ORCID

classification eess.SP

keywords 5GNRPUSCHURLLCCRCpredictionscatteringtransformevidentialdeeplearningHARQlatencyMCSconditioning

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 proposes that a neural network can predict the cyclic redundancy check (CRC) outcome of a 5G NR PUSCH transport block directly from the received I/Q resource grid and least-squares channel estimates, bypassing equalization, demodulation, and LDPC decoding. If the claim holds, a base station would know before the slot boundary which blocks fail, allowing it to schedule a retransmission in the very next uplink slot and reduce the effective HARQ round-trip time from 5 ms to 0.5 ms. The design combines a fixed scattering front-end with a compact CNN and an auxiliary MCS-index input; the best configuration reports 95% accuracy, 3.91% missed-detection rate, and 10.00% false-alarm rate on an out-of-distribution test set. An evidential extension using a conservative uncertainty rule further lowers the missed-detection rate to 0.76%.

What carries the argument

The load-bearing mechanism is a fixed first-order scattering front-end that converts the received signal and channel estimate into phase-invariant, multi-scale energy features. The Gabor realization applies J=2 scales and L=4 orientations as spatial depthwise convolutions; the FFT realization multiplies the input's two-dimensional Fourier transform by constant-quality-factor bandpass masks (Q=2) with center frequencies geometrically spaced from f_min=0.02 to f_max=0.4 and recovers the features by inverse transform and modulus. Both versions feed a compact CNN classifier with roughly 7,400 trainable parameters, and the MCS index enters as a learned embedding concatenated with the pooled features before the final sigmoid. The evidential extension replaces the sigmoid with Dirichlet concentration parameters, enabling a single-pass decomposition of total uncertainty into aleatoric and epistemic parts and a conservative decision rule for predicting CRC failure.

What would settle it

Feed the FFT+CNN+MCS predictor test slots in which a deep narrowband fade falls between DMRS subcarriers, and compare its missed-detection rate with that of the same model retrained on a two-DMRS pilot pattern; a sharp degradation in the single-DMRS version would show that the interpolation assumption, not the classifier, is the limiting component.

Watch

Extended reading notes

Core claim

The central discovery is that the post-decoding CRC verdict of a 5G NR PUSCH transport block is learnable from raw received symbols plus a least-squares channel estimate, without running the conventional equalization–demodulation–decoding chain. The paper shows that a fixed first-order scattering front-end, realized either with Gabor filters or with FFT-domain bandpass masks at geometrically spaced scales, supplies the multi-scale time-frequency energy features a compact CNN needs: replacing a pure CNN with either front-end raises accuracy from 82% to 90% and cuts the missed-detection rate from 19.20% to below 4.5%. Conditioning on the modulation and coding scheme (MCS) index lets the model adapt its decision boundary to the operating code rate, cutting the FFT variant's false-alarm rate from 26.21% to 10.00% while raising accuracy to 95%. The evidential output head, using a conservative rule that predicts failure when epistemic uncertainty exceeds a threshold, reaches 0.76% missed detections, and both front-ends execute within the slot processing budget on an edge-class GPU.

Load-bearing premise

The network must learn decodability from a single-DMRS least-squares channel estimate expanded across the full resource grid; if that expansion hides the fine-grained per-subcarrier SNR or frequency-selective fading structure that determines LDPC decodability, the predictor cannot recover it.

Editorial extensions

If this is right

A gNB can schedule a retransmission in the next uplink slot, cutting effective HARQ round-trip time from 5 ms to 0.5 ms for mini-slot URLLC.
Configured-grant blind repetitions (2–8×) can be replaced by triggered retransmissions, saving air-interface resources while preserving deterministic latency.
The same trained evidential predictor can serve both eMBB (standard rule) and URLLC (conservative rule) by switching decision thresholds at inference.
MCS conditioning makes one predictor span six code rates; the FFT front-end's real compute is 0.299 ms on a full 273-PRB grid and scales down with URLLC-sized allocations.
The decision thresholds (tau and tau_u) give operators deployment-time knobs to trade missed detections against unnecessary retransmissions.

Reading between the lines

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

A testable extension is to compare the single-DMRS expanded channel estimate with denser pilot configurations; if accuracy improves substantially, the interpolation step, not the network, is the current ceiling.
The same early-CRC logic should transfer to downlink reception or other CRC-protected channels, where the same timing pressure exists.
Because false alarms waste resources but missed detections are silent, operators could adapt the thresholds per quality-of-service flow rather than per network.
The 5 ms-to-0.5 ms latency claim assumes the predictor's own runtime fits between the mini-slot's end and the slot boundary; measuring end-to-end MAC-scheduler integration would confirm whether the theoretical gain survives protocol overhead.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Hey,

Quick take: this is a genuinely new application—predicting post-LDPC CRC outcome directly from the received I/Q grid and LS channel estimates, before decoding. The authors know the neural-receiver literature and borrow its input conventions cleanly. The scattering front-end plus MCS conditioning is a sensible combination, and the simulation results are internally consistent: both scattering models beat the plain CNN by a wide margin on MDR, and MCS conditioning roughly halves FAR for the FFT variant. The framing of MDR vs FAR in terms of URLLC retransmission cost is correct and useful. This is the strongest part of the paper.

The soft spots are real and mostly about evidence quality. There are no error bars or multiple seeds; the decision threshold and the epistemic threshold are tuned on the test set; and no code or data are released. I can't falsify the central claim, but I can't fully trust the specific numbers either.

The bigger problem is the latency/deployment claim. The paper's motivating scenario is a 7-symbol URLLC mini-slot, and the stated budget is ~142 us before the slot boundary. But the model is trained and evaluated only on a 13-symbol, 273-PRB PUSCH slot. Measured TensorRT real compute for FFT+CNN is 0.299 ms—more than twice the budget. The paper hand-waves that narrowband inference will be proportionally shorter, but that is not a measurement, and there is no evaluation on 7-symbol grids at all. So the 5ms-to-0.5ms HARQ RTT reduction is unsupported by the presented data. This is a load-bearing flaw for the deployment claim, not a nitpick.

One mitigating note: I think the stress-test's worry about single-DMRS LS interpolation is mostly a non-issue. The received data symbols themselves carry channel-state information, so the network has a redundant path to infer decodability. Not a fatal concern.

Bottom line: the core idea deserves a serious referee. It is a legitimate research direction with a coherent simulation study. What it needs is a revision that reports variance, releases the artifacts, and produces actual narrowband mini-slot latency and accuracy measurements. As it stands, I'd send it to review with major revision requested, not desk reject.

Referee Report

3 major / 5 minor

Summary. The paper proposes a neural CRC predictor for 5G NR PUSCH that takes the received I/Q resource grid and LS channel estimates, extracts first-order scattering features via either Gabor or FFT-based fixed front-ends, and classifies CRC pass/fail with a compact CNN, optionally conditioned on the MCS index. On an NVIDIA Aerial simulation with 30,000 training and 20,000 test slots, the scattering front-ends are reported to substantially outperform a pure CNN, MCS conditioning reduces the false-alarm rate of the scattering models, and the best operating point is FFT+CNN+MCS with 95% accuracy, 3.91% missed-detection rate, and 10.00% false-alarm rate. An evidential extension with a conservative uncertainty rule is shown to reduce MDR to 0.76% at 42.24% FAR. The paper claims this enables reducing effective HARQ round-trip time from 5 ms to 0.5 ms for URLLC mini-slots.

Significance. If the operational claims held, the paper would demonstrate a practically useful mechanism for early retransmission scheduling in URLLC. The empirical core is internally consistent: the comparison of scattering front-ends against a pure CNN uses large margins, the fixed front-ends add essentially no trainable parameters, the TensorRT compatibility is a concrete strength, and the evidential-uncertainty extension is a reasonable single-pass alternative to sampling-based methods. The main weakness is that the deployment-level claim is not supported by the measurements: all experiments use a full-bandwidth, 13-symbol slot, whereas the motivating scenario is a 7-symbol mini-slot with a 142 us processing budget, and the measured FFT+CNN real compute time of 0.299 ms already exceeds that budget. The lack of multi-seed confidence intervals also weakens the finer-grained comparisons between front-ends and the evidential threshold results. These issues are fixable with additional experiments or a qualified reframing.

major comments (3)

[Section 1.1 and Section 4.5] The central deployment claim—that the predictor reduces HARQ RTT from 5 ms to 0.5 ms by completing within the ~142 us left after a 7-symbol mini-slot—is not supported by any presented measurement. All experiments use a 273-PRB, 13-symbol PUSCH slot (Table 2), and the measured TensorRT real compute for FFT+CNN is 0.299 ms (Section 4.5), which exceeds the 0.142 ms slot-boundary budget. The statement that inference time is 'expected to be proportionally shorter' for narrowband URLLC allocations is an extrapolation; no narrowband latency measurement and no mini-slot accuracy evaluation are reported. Because the model is trained and evaluated on 13-symbol grids, its behavior on shorter or narrower grids is untested. Please either measure the actual mini-slot configuration or restrict the abstract's and conclusion's RTT-reduction claims to configurations consistent with the experiments.
[Section 4.2 and Table 3] All headline numbers are single point estimates from one train/test split, with no seeds, confidence intervals, or significance tests. The front-end comparisons used to justify design choices—for example, FFT+CNN versus Gabor+CNN with MCS conditioning (MDR 3.91% versus 2.90%; FAR 10.00% versus 18.28%)—include differences of a few percent that could be within run-to-run variability. In addition, the evidential uncertainty threshold tau_u = 0.025 in Table 3 appears to be chosen on the test set; if so, the reported 0.76% MDR is an optimistic in-sample selection result. Please report means and standard deviations over multiple seeds and use a validation-based procedure for threshold selection.
[Section 4.6] The conservative decision rule of Eq. (29) effectively reclassifies all high-epistemic-uncertainty samples as failures; at tau_u = 0.025, 19.3% of test samples are flagged and FAR rises from 10.00% to 42.24%. The paper justifies this as acceptable by comparing expected retransmission count with blind repetition, but this ignores that each false alarm consumes a scheduling opportunity and adds latency or jitter for a TB that would have succeeded. A proper evaluation should report the MDR-versus-FAR operating curve over tau_u and compare at matched FAR, or simulate the resulting HARQ process, before claiming that the evidential rule provides the best safety margin for URLLC.

minor comments (5)

[Section 3.2] The text says 'The modulus in (9)' but the modulus operation is defined in Eq. (11); Eq. (9) is the concatenation of received symbols and channel estimates.
[Fig. 3] The bar chart labels overlap the bars, making the figure difficult to read; the normalized-latency series duplicates information from Section 4.5 and should be shown in a separate panel.
[Table 2 and Section 4.1] The test set is described as out-of-distribution in delay spread and UE speed, but Table 2 lists the same delay-spread range for training and inference; please specify exactly which delay-spread/UE-speed combinations are held out and whether SNR or channel model realizations differ.
[References] Several references have formatting errors (e.g., [2] lists 'Huttunen, J. M. J. , M.'), and the manuscript does not include a data/code availability statement, which would aid reproducibility of the numerical claims.
[Section 2.4] The statement that a false alarm 'wastes resources, but far less than blind repetition' assumes that one false alarm costs exactly one retransmission; under configured-grant or retransmission-limit operation the cost may be higher, so the claim should be qualified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core result is a supervised held-out evaluation with external, non-self references and an acknowledged mathematical equivalence between the two front-ends.

full rationale

I find no circular step that can be substantiated from the paper's text. The central claim is an empirical supervised-learning result: the network is trained on the received resource grid, LS channel estimates, and MCS index against ground-truth CRC labels produced by the conventional LDPC decoder (Eqs. 4-5), then evaluated on a held-out out-of-distribution test set described in Table 2. The labels are not derived from the network's own outputs, and the reported accuracy, MDR, and FAR are measured on test data rather than obtained from fitted constants. The FFT and Gabor front-ends are connected by the convolution theorem (Eqs. 13-16), and the paper explicitly states this equivalence, so no ansatz is smuggled in via citation. References are to external prior work (DeepRx, scattering transforms, evidential learning, NVIDIA Aerial) and are not self-citations by the author; no uniqueness theorem from the author's own prior work is invoked to force a choice. The epistemic uncertainty threshold tau_u = 0.025 in Sec. 4.6 is presented as a deployment-time knob (Eq. 29) rather than as a parameter fitted to the test labels and then renamed a prediction; the paper does not state that the threshold was selected from test-set outcomes, so under the hard rules I cannot exhibit a circular reduction. The closest concern, that the conservative decision rule's low MDR may depend on threshold choice, is a threshold-selection or data-snooping matter, not a circularity of the derivation chain, and the paper does not provide enough text to establish it as a fitted-input-called-prediction step.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claim rests primarily on a set of hand-chosen hyperparameters and domain assumptions about the simulator and channel model. No new physical entities are introduced. The most consequential free parameters are the epistemic uncertainty threshold tau_u, which drives the headline 0.76% MDR, and the scattering frequency range, which defines the feature representation. The axioms are standard for simulation-based wireless ML research, but the single-DMRS LS interpolation assumption and the simulator-fidelity assumption are load-bearing and unverified.

free parameters (7)

Scattering scales and orientations (J=2, L=4) = J=2, L=4
Hand-chosen filter bank configuration giving 8 first-order coefficients per input channel; no ablation or sensitivity analysis is reported (Sections 3.2, 4.4).
Scattering frequency range f_min, f_max = 0.02, 0.4
Hand-selected normalized spatial frequencies justified qualitatively as covering slow channel envelope and fine spectral nulls; no optimization (Section 3.3).
Constant-Q bandwidth factor b_j/f_j = 0.5 (Q=2)
Hand-chosen bandwidth scaling for the FFT-based masks; no sensitivity study is provided (Section 3.3).
Decision threshold tau = 0.5 (default)
Fixed classification threshold in Eq. (6); the paper notes it can be tuned but does not systematically analyze the operating curve (Sections 2.3, 4.3).
Epistemic uncertainty threshold tau_u = 0.025
Chosen to produce the reported 0.76% MDR in Table 3; the paper does not describe a validation-based selection, indicating a test-set-tuned operating point (Section 4.6, Eq. 29).
Evidential KL weight lambda = 0.01
Regularization strength in Eq. (23); no ablation over lambda is reported (Section 4.6).
CNN hidden dimension = 8 channels baseline, 16 in scaled-up variant
Architecture hyperparameter; only one scaled-up variant is tested, with no broader sweep (Section 4.4).

assumptions (5)

domain assumption 5G NR PUSCH signal model: received signal is the transmitted symbol multiplied by a frequency-selective channel response plus additive complex Gaussian noise (Eq. 1).
This standard linear model underlies the entire simulation and feature design; the paper uses TDL/CDL channels and AWGN in NVIDIA Aerial (Sections 2.1, 4.1).
domain assumption LS channel estimates from a single DMRS symbol, expanded to the full resource grid, preserve the per-subcarrier SNR and channel structure needed for CRC prediction.
The input representation u in Eq. (9) is the concatenation of received symbols and expanded LS estimates; if interpolation loses frequency-selective information, the predictor cannot recover it (Section 3.1).
domain assumption The NVIDIA Aerial simulator correctly implements 5G NR PHY processing including LDPC decoding and CRC, making the ground-truth labels trustworthy.
All training labels and test labels come from the simulator's conventional decoder; there is no independent hardware or over-the-air validation (Section 4.1).
ad hoc to paper First-order scattering coefficients are sufficient; higher-order scattering is not needed for this task.
The paper retains only first-order scattering features and provides no ablation showing that higher-order coefficients would not improve performance (Sections 1.2, 3.2).
domain assumption The out-of-distribution test set (UE speed up to 140 kmph, delay spread 0-300 ns) is representative of real deployment channels.
The generalization claim rests on these held-out parameter ranges, but the test data still come from the same TDL/CDL simulator family as training (Section 4.1).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Neural CRC Prediction for 5G NR URLLC." pith.science (2026). https://pith.science/paper/SY4RCRRV

@misc{pith2026260806230,
  author       = {Pith},
  title        = {Pith review of: Neural CRC Prediction for 5G NR URLLC},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SY4RCRRV}},
  note         = {Machine review of arXiv:2608.06230}
}

read the original abstract

We propose a neural cyclic redundancy check (CRC) predictor for the 5G New Radio (5G NR) physical uplink shared channel (PUSCH) that enables early link-adaptation decisions for Ultra-Reliable Low-Latency Communications (URLLC). The predictor combines a lightweight convolutional neural network (CNN) with a fixed front-end that extracts multi-scale time-frequency energy features from the received signal and least-squares channel estimates. We investigate two complementary front-end realizations - a wavelet scattering front-end built from fixed Gabor filters, and an FFT-based scattering front-end that applies bandpass masks in the frequency domain with geometric scale spacing. Drawing on neural-receiver design principles, the predictor estimates the post-decoding CRC outcome directly from the received resource grid, bypassing the conventional equalization and decoding chain. We further introduce the modulation and coding scheme (MCS) index as an auxiliary conditioning input that adapts the decision boundary to the operating code rate. Experiments on a multi-MCS 5G NR PUSCH dataset show that the hybrid scattering predictors substantially outperform a pure CNN baseline, with MCS conditioning further improving reliability across varying channel conditions. Both front-ends are compatible with GPU-accelerated inference and are lightweight enough to be deployed within a real-time baseband pipeline. We further demonstrate an evidential deep learning extension that quantifies epistemic uncertainty in a single forward pass using a conservative decision rule.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 7 canonical work pages

[1]

M., and Hassanieh, H.: Ultra-Reliable Low-Latency in 5G: A Close Reality or a Distant Goal?

Maghsoudnia, A., Vlad, E., Gong, A., Dumitriu, D. M., and Hassanieh, H.: Ultra-Reliable Low-Latency in 5G: A Close Reality or a Distant Goal?. In: Proceedings of the 23rd ACM Workshop on Hot Topics in Networks (HotNets ’24), pp. 111–120, Irvine, CA, USA, Nov

work page
[2]

D., and Huttunen, J

Honkala, J., Korpi. D., and Huttunen, J. M. J. , M.: DeepRx: Fully Convolutional Deep Learning Receiver . In: IEEE Transactions on Wireless Communications, 2021 . https://doi.org/10.48550/arXiv.2005.01494

arXiv doi:10.48550/arxiv.2005.01494 2021
[3]

Model-Driven Deep Learning for MIMO Detection

He, H., Wen, C.-K., Jin, S., Li, G. Y.: Model-Driven Deep Learning for MIMO Detection. In: IEEE Transactions on Signal Processing, 2020 . https://doi.org/10.48550/arXiv.1907.09439

work page Pith review arXiv doi:10.48550/arxiv.1907.09439 2020
[4]

Learning to Detect

Samuel, N., Diskin, T., Wiesel, A.: Learning to Detect. In: IEEE Transactions on Signal Processing, 2019. https://doi.org/10.48550/arXiv.1805.07631

work page Pith review arXiv doi:10.48550/arxiv.1805.07631 2019
[5]

A Neural Receiver for 5G NR Multi-user MIMO

Cammerer, S., Aoudia, A. F., Hoydi s, J., Oeldemann, A., Roessler, A., Mayer, T., and Keller, A.: A Neural Receiver for 5G NR Multi-user MIMO. In: Proceedings of the IEEE Globecom Workshops, Mar. 2023 , pp. 329 –334. https://doi.org/10.48550/arXiv.2312.02601

work page Pith review arXiv doi:10.48550/arxiv.2312.02601 2023
[6]

, Cammerer, S., Aoudia, F

Wiesmayr, R. , Cammerer, S., Aoudia, F. A., Hoydis, J., Zakrzewski, J., and Keller, A.: Design of a Standard-Compliant Real-Time Neural Receiver for 5G NR. In: 2025 IEEE In- ternational Conference on Machine Learning for Communication and Networking (ICMLCN), Barcelona, Spain, 2025. https://doi.org/10.48550/arXiv.2409.02912 20

arXiv doi:10.48550/arxiv.2409.02912 2025
[7]

In: IEEE Transactions on Wireless Communications, vol

Gupta, A., Dizdar, O., Chen, Y., Emre Kadan, F., Sattarzadeh, A., and Wang, S., : Neu- romorphicRx: From Neural to Spiking Receiver . In: IEEE Transactions on Wireless Communications, vol. 25, pp. 10161 -10174, 2026 . https://doi.org/10.48550/arXiv.2512.05246

work page doi:10.48550/arxiv.2512.05246 2026
[8]

In: IEEE Transac- tions on Pattern Analysis and Machine Intelligence, vol

Bruna, J., and Mallat, S.: Invariant Scattering Convolution Networks. In: IEEE Transac- tions on Pattern Analysis and Machine Intelligence, vol. 35, no. 8, pp. 1872 –1886, Aug

work page

Show all 13 references

[9]

In: IEEE Transactions on Signal Pro- cessing, vol

Andén, J., and Mallat, S.: Deep Scattering Spectrum. In: IEEE Transactions on Signal Pro- cessing, vol. 62, no. 16, pp. 4114 –4128, Aug. 2014 . https://doi.org/10.1109/TSP.2014.2326991

2014
[10]

In: Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), 2018

Sensoy, M., Kaplan, L., and Kandemir, M.: Evidential Deep Learning to Quantify Classi- fication Uncertainty. In: Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), 2018. https://doi.org/10.48550/arXiv.1806.01768

2018 doi arXiv
[11]

NVIDIA Aerial CUDA -Accelerated RAN , https://github.com/NVIDIA/aerial-cuda- accelerated-ran
[2013]

https://doi.org/10.1109/TPAMI.2012.230

2012 doi
[2024]

https://doi.org/10.1145/3696348.3696862

Pith tools

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

[1] [1]

M., and Hassanieh, H.: Ultra-Reliable Low-Latency in 5G: A Close Reality or a Distant Goal?

Maghsoudnia, A., Vlad, E., Gong, A., Dumitriu, D. M., and Hassanieh, H.: Ultra-Reliable Low-Latency in 5G: A Close Reality or a Distant Goal?. In: Proceedings of the 23rd ACM Workshop on Hot Topics in Networks (HotNets ’24), pp. 111–120, Irvine, CA, USA, Nov

work page

[2] [2]

D., and Huttunen, J

Honkala, J., Korpi. D., and Huttunen, J. M. J. , M.: DeepRx: Fully Convolutional Deep Learning Receiver . In: IEEE Transactions on Wireless Communications, 2021 . https://doi.org/10.48550/arXiv.2005.01494

arXiv doi:10.48550/arxiv.2005.01494 2021

[3] [3]

Model-Driven Deep Learning for MIMO Detection

He, H., Wen, C.-K., Jin, S., Li, G. Y.: Model-Driven Deep Learning for MIMO Detection. In: IEEE Transactions on Signal Processing, 2020 . https://doi.org/10.48550/arXiv.1907.09439

work page Pith review arXiv doi:10.48550/arxiv.1907.09439 2020

[4] [4]

Learning to Detect

Samuel, N., Diskin, T., Wiesel, A.: Learning to Detect. In: IEEE Transactions on Signal Processing, 2019. https://doi.org/10.48550/arXiv.1805.07631

work page Pith review arXiv doi:10.48550/arxiv.1805.07631 2019

[5] [5]

A Neural Receiver for 5G NR Multi-user MIMO

Cammerer, S., Aoudia, A. F., Hoydi s, J., Oeldemann, A., Roessler, A., Mayer, T., and Keller, A.: A Neural Receiver for 5G NR Multi-user MIMO. In: Proceedings of the IEEE Globecom Workshops, Mar. 2023 , pp. 329 –334. https://doi.org/10.48550/arXiv.2312.02601

work page Pith review arXiv doi:10.48550/arxiv.2312.02601 2023

[6] [6]

, Cammerer, S., Aoudia, F

Wiesmayr, R. , Cammerer, S., Aoudia, F. A., Hoydis, J., Zakrzewski, J., and Keller, A.: Design of a Standard-Compliant Real-Time Neural Receiver for 5G NR. In: 2025 IEEE In- ternational Conference on Machine Learning for Communication and Networking (ICMLCN), Barcelona, Spain, 2025. https://doi.org/10.48550/arXiv.2409.02912 20

arXiv doi:10.48550/arxiv.2409.02912 2025

[7] [7]

In: IEEE Transactions on Wireless Communications, vol

Gupta, A., Dizdar, O., Chen, Y., Emre Kadan, F., Sattarzadeh, A., and Wang, S., : Neu- romorphicRx: From Neural to Spiking Receiver . In: IEEE Transactions on Wireless Communications, vol. 25, pp. 10161 -10174, 2026 . https://doi.org/10.48550/arXiv.2512.05246

work page doi:10.48550/arxiv.2512.05246 2026

[8] [8]

In: IEEE Transac- tions on Pattern Analysis and Machine Intelligence, vol

Bruna, J., and Mallat, S.: Invariant Scattering Convolution Networks. In: IEEE Transac- tions on Pattern Analysis and Machine Intelligence, vol. 35, no. 8, pp. 1872 –1886, Aug

work page