Pith. sign in

REVIEW 3 major objections 6 minor 53 references

A Semantic Approach to Successive Interference Cancellation for Multiple Access Networks

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Successive interference cancellation for multi-user channels can be moved from the digital symbol domain into the semantic word-embedding domain, where decoded text from one user helps decode the next.

desk verdict Real extension of DeepSC to the MAC with semantic side-information SIC, but the side-information gain is only shown on deliberately correlated SNLI texts, and the traditional baselines lack SIC. read the letter →

arxiv 2501.10926 v1 pith:ANKCUHTI submitted 2025-01-19 cs.IT math.IT

Mingxiao Li , Kaiming Shen , Shuguang Cui This is my paper · ORCID
classification cs.ITmath.IT
keywords semanticcommunicationsuccessiveinterferencecancellationmultipleaccesschannelDeepSCTransformerwordembeddingsideinformationpartialretraining
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

The paper proposes a successive interference cancellation (SIC) scheme that works in the semantic domain, extending deep-learning semantic communication from a single point-to-point link to a multi-user multiple access channel (MAC). In classic SIC, once a user's message is decoded, only their signal is subtracted from the received waveform; here, the decoded text itself is kept and used as side information to help decode the next user's text, exploiting semantic correlations between users' messages. The paper also contributes a pretraining and partial retraining scheme that lets new users join an existing MAC without retraining all of the old encoders and decoders from scratch. The reported experiments claim that the side-information variants achieve higher BERT-similarity and BLEU scores than full retraining without side information, and far outperform classic digital coding baselines using 64-QAM.

What carries the argument

The load-bearing mechanism is a neural module called the integrated feature generator (IFG). For user $i$'s sentence, it takes the current compressed word-semantic vector $\hat{r}^i_j$ together with the recovered compressed vectors of previously decoded users $\tilde{r}^1_j, \ldots, \tilde{r}^{i-1}_j$ and produces $g^i_j = \text{ReLU}(\theta_i(\pi_i(\hat{r}^i_j), \Omega^i_1(\tilde{r}^1_j), \ldots, \Omega^{i-1}_i(\tilde{r}^{i-1}_j)) + \hat{r}^i_j)$, which the autoencoder decoder converts into the word-semantic vector used to generate the next text. This fusion is what changes SIC from a pure subtraction operation into a side-information operation. Around this, the training scheme uses a joint cross-entropy loss, a per-user pretraining stage, and a partial retraining algorithm that freezes the existing users' encoders and decoders when new users join, so only the new users' networks and the shared decoder are updated.

What would settle it

Run the same system on a MAC whose users transmit semantically unrelated texts, for example sentences sampled independently from different topics so there is no entailment-style pairing between users, and compare the with-side-information and without-side-information variants under BERT similarity and BLEU. If the side-information variants do not beat the no-side-information baselines, the claimed benefit rests entirely on the dataset's inter-sentence correlations and does not hold for general user traffic.

Watch

Extended reading notes

Core claim

The paper's central claim is that SIC for a multiple access channel can be performed in the semantic domain rather than only in the digital symbol domain. After the strongest user's text is recovered, the usual SIC step is kept, namely reconstructing and subtracting that user's symbols from the received signal, but the recovered text itself is not discarded. Its compressed word-semantic vectors are passed through an integrated feature generator together with the next user's noisy features, and the combined feature is decoded into the next text. The paper argues that because natural-language messages from different users are often semantically related, this side information improves semantic similarity and BLEU scores, and that a partial retraining scheme which freezes old users' networks and trains only the new users' networks preserves most of the benefit while reducing per-iteration training time by more than 30 percent.

Load-bearing premise

The gain from side information depends on users' texts being semantically related, so a decoded sentence from one user genuinely narrows the possibilities for another user's sentence; without that correlation the side information may add nothing or even hurt.

Editorial extensions

If this is right

  • In a K-user MAC, decoding can proceed in order of descending received power, and each successfully decoded text is fed into the decoding of the next text rather than discarded.
  • Adding a new user to a trained MAC can be handled by partial retraining that keeps existing encoders and decoders fixed, reducing per-iteration training time by more than 30 percent relative to full retraining with side information.
  • The reported results imply side information is most valuable in low-SINR and fading conditions, where partial retraining with side information outperforms full retraining without it.
  • Because the method is built on SIC, the same semantic-side-information idea can be carried to broadcast and interference channels whenever SIC applies, although capacity achievability is not guaranteed there.

Reading between the lines

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

  • A testable extension is to modulate the side-information branch of the IFG by an estimated inter-user semantic correlation, letting the network ignore side information when users' messages are unrelated; this would reveal how much of the reported gain is dataset-specific.
  • The subtract-then-reuse-recovered-content recipe should transfer to image or audio semantic MACs, replacing BERT/BLEU with SSIM or SDR, which the authors themselves list as future work.
  • Freezing old users' networks during partial retraining implies an incremental onboarding protocol for IoT devices that speak semantic protocols: new devices can enter the MAC without reconfiguring existing ones, which is a deployment property beyond the tested performance metrics.
  • If semantic SIC holds up, the effective multiuser limit of such systems depends on inter-user semantic redundancy as much as on power and bandwidth, suggesting a semantic capacity region notion that the paper does not formalize.
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. This paper proposes a deep learning-enabled semantic communication system for uplink multiple access channels (MAC), extending the single-user DeepSC framework to K users. The approach applies successive interference cancellation (SIC) in the semantic domain: after decoding each user's text, the decoded text is re-encoded and its signal is removed from the received mixture; in addition, an integrated feature generator (IFG) uses previously decoded texts as side information to improve decoding of subsequent users. To handle dynamic user addition, the paper also proposes a pretraining scheme and a partial retraining scheme that updates only the newly added users' networks while reusing existing users' encoders/decoders. Experiments on the SNLI corpus compare the proposed methods with ablated baselines (with/without side information, full/partial retraining) and with traditional Huffman+LDPC and Huffman+Polar schemes, reporting semantic similarity and BLEU scores across several channel conditions.

Significance. If the claims hold, the paper provides a meaningful extension of semantic communication to multi-user networks and introduces a practical retraining mechanism for dynamic user sets. The internal ablations show consistent gains from side information and reduced training time from partial retraining, and the open-source code is a plus. However, the significance is somewhat limited by the evaluation: the traditional baselines do not use any interference cancellation, so the comparison does not isolate the value of the semantic SIC; and the side-information benefit is demonstrated only on SNLI, which was selected for its strong inter-sentential correlations, leaving the behavior on uncorrelated sources untested.

major comments (3)
  1. [Section VI-A] The traditional baselines (Huffman+LDPC and Huffman+Polar with 64-QAM) are described without any interference cancellation or multi-user detection, whereas the proposed method relies on successive interference cancellation. As a result, the performance differences in Figs. 9–12 could be attributed to SIC itself rather than its semantic aspect. Please include a baseline that performs conventional symbol-level SIC with the same channel coding, or justify why the current comparison is appropriate for the claimed advantage over existing benchmark methods.
  2. [Section VI-A] The side-information benefit is evaluated only on the SNLI dataset, which the authors selected because it exhibits much stronger inter-sentential connections than other corpora. The mechanism in Eqs. (34)–(35) is thus tested exclusively in a high-correlation regime. No experiment varies the degree of semantic correlation between user texts (e.g., using random sentence pairings or a corpus such as Europarl). Without such an experiment, it is unclear whether the 'semantic SIC' advantage persists when users transmit unrelated content, which is a common scenario in multiple access networks. Please add such experiments or explicitly qualify the claims to correlated sources.
  3. [Section V-C, Eq. (39) and Algorithm 2] The tradeoff factors τ_i in the loss function (39) are not specified; it is unclear how they are set or whether they are tuned. Since the partial retraining performance is a central contribution, please provide the values used or a sensitivity analysis to show that the results are not sensitive to these factors.
minor comments (6)
  1. [Section III, Eq. (19)] The zero-padding procedure in Eq. (19) pads each sentence to a fixed length N; however, the paper does not specify what happens when a sentence exceeds N words (truncation?) or how the positional encoding handles the padded dummy words. Clarify this in the text.
  2. [Section V-C and Algorithm 4] The description of the partial retraining decoding order is inconsistent: Section V-C states that after recovering the new user's message, the old users in group G2 are decoded 'as if the new user does not exist', but Algorithm 4 decodes all old users in a single step after canceling the new users' signals. Please reconcile the algorithm with the narrative.
  3. [Section VI-A] The statement 'we allocate 18 symbols for each word' is not connected to the encoding parameters: in the system model, each word is compressed to a c-dimensional vector and then mapped to c/2 complex symbols. Please explain how 18 symbols per word relates to c and the total block length M.
  4. [Section IV-B] The paper calls the method 'semantic SIC', but the actual interference cancellation in Eqs. (30) and (32) is performed in the complex symbol domain after re-encoding; only the side-information generation (Eq. (35)) is in the semantic embedding domain. Consider clarifying this terminology.
  5. [Section VI-B, Fig. 11] The text contains a typo: 'acheive' should be 'achieve' in the description of the Rayleigh fading BLEU results.
  6. [References] Reference [47] (SNLI) lists the arXiv identifier 2201.01389, which duplicates reference [13]; the correct identifier for the SNLI corpus is 1508.05326. Please correct this citation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the semantic-SIC claims are supported by independent training and evaluation, not by self-citation or construction.

full rationale

The paper's derivation chain is empirical and self-contained. The semantic encoding and decoding modules follow the DeepSC framework, which is cited as prior work rather than derived from the paper's own claims. The central contribution, semantic SIC with side information, is implemented through the IFG mechanism in Eqs. (34)-(35), and its benefit is evaluated by comparing full and partial retraining with side information against the same procedures without side information, as well as against traditional Huffman+LDPC and Huffman+Polar baselines. The loss functions Ljoint (Eq. 38) and LFP (Eq. 39) are standard supervised objectives; no trained parameter is relabeled as a prediction, and no target quantity is defined in terms of the method's output. The choice of the SNLI dataset because of strong inter-sentential connections is a scope limitation: it means the side-information gain may not generalize to uncorrelated user texts, but it does not make the comparison circular. Citations to the authors' own prior work are peripheral and not load-bearing, and no uniqueness theorem or ansatz is imported from self-citations. The inserted revision remarks in the manuscript are descriptive and do not assert a circular step or omitted proof. Overall, the central empirical claims rest on independent benchmarks and held-out test evaluation, so the paper exhibits no significant circularity.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The central claim depends on the chosen architecture and training hyperparameters, as well as the domain assumption that users share a Gaussian MAC and that their texts are semantically related. The IFG is the only new architectural entity; it has no independent evidence outside the paper's experiments.

free parameters (5)
  • compression dimension c = not explicitly stated; 18 channel symbols per word implies c=36
    chosen by hand to balance rate and distortion; directly determines the number of channel uses per sentence.
  • embedding dimension d = 128
    stated in the footnote as the default word embedding dimension.
  • max sentence length N = not explicitly stated; preprocessing limits sentences to 4 to 20 words
    zero padding to a common length N is required to form fixed-length vectors; N is a design choice.
  • number of Transformer layers = 4
    stated as a 4-layer Transformer; architecture choice.
  • tradeoff factor tau_i in loss (39) = not specified
    introduced to balance old and new user losses, but no values are given.
assumptions (5)
  • domain assumption Channel model: y_m = sum_i h_i x^i_m + z_m with i.i.d. Gaussian noise (Eq. 4)
    The entire system model assumes linear superposition of user signals and additive white Gaussian noise; this is standard for MAC analysis but is an assumption about the physical channel.
  • domain assumption Power constraint on each user's transmit sequence (Eq. 3)
    Each user normalizes its transmitted symbol vector to average power P_i; this is a physical constraint.
  • domain assumption Decoding order is fixed by decreasing received power (Eq. 1)
    SIC requires a decoding order; the paper assumes users can be re-indexed so that P_1|h_1|^2 >= ... >= P_K|h_K|^2.
  • standard math The Transformer, BERT embeddings, and autoencoders are trainable with standard backpropagation and Adam
    The training procedure relies on standard deep learning assumptions: differentiability, stochastic optimization, and generalization from a training split.
  • ad hoc to paper Semantic correlation between users' texts is present and exploitable
    The side-information contribution depends on this; the paper deliberately selects SNLI, a dataset with strong inter-sentential connections, to make this assumption hold.
invented entities (1)
  • Integrated Feature Generator (IFG) network
    purpose: Fuses the current user's compressed semantic vectors with side information from previously decoded users to improve decoding
    A new neural network component introduced by this paper; its effectiveness is only demonstrated on the chosen dataset and it has no externally validated property.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Semantic Approach to Successive Interference Cancellation for Multiple Access Networks." pith.science (2026). https://pith.science/paper/ANKCUHTI

@misc{pith2026250110926,
  author       = {Pith},
  title        = {Pith review of: A Semantic Approach to Successive Interference Cancellation for Multiple Access Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ANKCUHTI}},
  note         = {Machine review of arXiv:2501.10926}
}
read the original abstract

Differing from the conventional communication system paradigm that models information source as a sequence of (i.i.d. or stationary) random variables, the semantic approach aims at extracting and sending the high-level features of the content deeply contained in the source, thereby breaking the performance limits from the statistical information theory. As a pioneering work in this area, the deep learning-enabled semantic communication (DeepSC) constitutes a novel algorithmic framework based on the transformer--which is a deep learning tool widely used to process text numerically. The main goal of this work is to extend the DeepSC approach from the point-to-point link to the multi-user multiple access channel (MAC). The inter-user interference has long been identified as the bottleneck of the MAC. In the classic information theory, the successive interference cancellation (SIC) scheme is a common way to mitigate interference and achieve the channel capacity. Our main contribution is to incorporate the SIC scheme into the DeepSC. As opposed to the traditional SIC that removes interference in the digital symbol domain, the proposed semantic SIC works in the domain of the semantic word embedding vectors. Furthermore, to enhance the training efficiency, we propose a pretraining scheme and a partial retraining scheme that quickly adjust the neural network parameters when new users are added to the MAC. We also modify the existing loss function to facilitate training. Finally, we present numerical experiments to demonstrate the advantage of the proposed semantic approach as compared to the existing benchmark methods.

Figures

Figures reproduced from arXiv: 2501.10926 by the authors.

Figure 1
Figure 1. Semantic communications in a 3-user MAC. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The paradigm of the proposed semantic decoder for the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. The paradigm of the proposed semantic encoder. The da [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: The structure of semantic decoder. The dashed boxes r [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Paradigm of the integrated feature generator [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Phase I of the adapted algorithm for one-user additio [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Phase II of the adapted algorithm for one-user additi [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Loss Comparison between the pretraining scheme and t [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Minimal semantic similarity across 2+1 users in the AW [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 10
Figure 10. Figure 10: Minimal BLEU score across 2+1 users in the AWGN case u [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: The minimal BLEU score for 2+1 users in the Rayleigh fa [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: The minimal semantic similarity for 3+2 users of in t [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 48 canonical work pages

  1. [1]

    A mathematical theory of communication,

    C. E. Shannon, “A mathematical theory of communication, ” Bell Syst. Tech. J., vol. 27, no. 3, pp. 379–423, Mar, 1948

    work page 1948

  2. [2]

    Nine challenges in artificial intell igence and wireless communications for 6G,

    W. Tong and G. Y . Li, “Nine challenges in artificial intell igence and wireless communications for 6G,” IEEE Wireless Commun. , vol. 29, no. 4, pp. 140–145, 2022

    work page 2022

  3. [3]

    Semantics-empowered communications: A tutorial-cum-su rvey,

    Z. Lu, R. Li, K. Lu, X. Chen, E. Hossain, Z. Zhao, and H. Zhan g, “Semantics-empowered communications: A tutorial-cum-su rvey,” IEEE Commun. Surveys Tuts. , 2023

    work page 2023

  4. [4]

    On the road to 6G: Visions, requirements, key technologies, and testbeds,

    C.-X. Wang, X. Y ou, X. Gao, X. Zhu, Z. Li, C. Zhang, H. Wang, Y . Huang, Y . Chen, H. Haas et al. , “On the road to 6G: Visions, requirements, key technologies, and testbeds,” IEEE Commun. Surveys Tuts., vol. 25, no. 2, pp. 905–974, 2023

    work page 2023

  5. [5]

    Semantic communication systems for s peech transmission,

    Z. Weng and Z. Qin, “Semantic communication systems for s peech transmission,” IEEE J. Sel. Areas Commun. , vol. 39, no. 8, pp. 2434– 2444, 2021

    2021

  6. [6]

    Autoencoder- based semantic communication systems with relay channels,

    X. Luo, B. Yin, Z. Chen, B. Xia, and J. Wang, “Autoencoder- based semantic communication systems with relay channels,” in Proc. IEEE Int. Conf. Commun. W orkshops (ICC W orkshops) , 2022, pp. 711–716

    work page 2022

  7. [7]

    Preserving qual ity of information by using semantic relationships,

    P . Basu, J. Bao, M. Dean, and J. Hendler, “Preserving qual ity of information by using semantic relationships,” Pervasive Mobile Comput., vol. 11, pp. 188–202, April, 2014

    work page 2014

  8. [8]

    Le ss data, more knowledge: Building next generation semantic communi cation networks,

    C. Chaccour, W. Saad, M. Debbah, Z. Han, and H. V . Poor, “Le ss data, more knowledge: Building next generation semantic communi cation networks,” IEEE Commun. Surveys Tuts. , 2024

    work page 2024

Show all 53 references
  1. [9]

    Energy efficient semantic communication over wireless networks with rate splitting,

    Z. Y ang, M. Chen, Z. Zhang, and C. Huang, “Energy efficient semantic communication over wireless networks with rate splitting, ” IEEE J. Sel. Areas Commun. , vol. 41, no. 5, pp. 1484–1495, 2023

    2023

  2. [10]

    DR L-driven dynamic resource allocation for task-oriented semantic co mmunication,

    H. Zhang, H. Wang, Y . Li, K. Long, and A. Nallanathan, “DR L-driven dynamic resource allocation for task-oriented semantic co mmunication,” IEEE Trans. Wireless Commun. , vol. 71, no. 7, pp. 3992–4004, 2023

    2023

  3. [11]

    Multimodal se mantic communication accelerated bidirectional caching for 6G ME C,

    C. Wang, X. Y u, L. Xu, Z. Wang, and W. Wang, “Multimodal se mantic communication accelerated bidirectional caching for 6G ME C,” Future Gener . Comput. Syst., vol. 140, pp. 225–237, 2023

    2023

  4. [12]

    Crossword: A semantic approach to text compression via masking,

    M. Li, R. Jin, L. Xiang, K. Shen, and S. Cui, “Crossword: A semantic approach to text compression via masking,” in Proc. IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP) , April 2024, pp. 9171–9175

    2024

  5. [14]

    Deep learning e nabled semantic communication systems,

    H. Xie, Z. Qin, G. Y . Li, and B.-H. Juang, “Deep learning e nabled semantic communication systems,” IEEE Trans. Signal Process., vol. 69, pp. 2663–2675, 2021

    2021

  6. [15]

    Robust semantic communications against semantic noise,

    Q. Hu, G. Zhang, Z. Qin, Y . Cai, G. Y u, and G. Y . Li, “Robust semantic communications against semantic noise,” in 2022 IEEE 96th V eh. Technol. Conf. (VTC Fall), Sep, 2022

    2022

  7. [16]

    Wireless deep video semantic transmission,

    S. Wang, J. Dai, Z. Liang, K. Niu, Z. Si, C. Dong, X. Qin, an d P . Zhang, “Wireless deep video semantic transmission,” IEEE J. Select. Areas Commun., vol. 41, no. 1, pp. 214 – 229, Jan 2022

    2022

  8. [17]

    YOLO-based semantic communication with generative AI-ai ded re- source allocation for digital twins construction,

    B. Du, H. Du, H. Liu, D. Niyato, P . Xin, J. Y u, M. Qi, and Y . T ang, “YOLO-based semantic communication with generative AI-ai ded re- source allocation for digital twins construction,” IEEE Internet Things J., 2023

    2023

  9. [18]

    Netwo rking architecture and key supporting technologies for human dig ital twin in personalized healthcare: a comprehensive survey,

    J. Chen, C. Yi, S. D. Okegbile, J. Cai, and X. Shen, “Netwo rking architecture and key supporting technologies for human dig ital twin in personalized healthcare: a comprehensive survey,” IEEE Commun. Surveys Tuts., vol. 26, no. 1, pp. 706–746, 2023

    2023

  10. [19]

    Metaverse commu- nications, networking, security, and applications: Resea rch issues, state- of-the-art, and future directions,

    M. Ali, F. Naeem, G. Kaddoum, and E. Hossain, “Metaverse commu- nications, networking, security, and applications: Resea rch issues, state- of-the-art, and future directions,” IEEE Commun. Surveys Tuts. , 2023

    2023

  11. [20]

    S emantic- aware sensing information transmission for metaverse: A co ntest the- oretic approach,

    J. Wang, H. Du, Z. Tian, D. Niyato, J. Kang, and X. Shen, “S emantic- aware sensing information transmission for metaverse: A co ntest the- oretic approach,” IEEE Trans. Wireless Commun. , vol. 22, no. 8, pp. 5214–5228, 2023

    2023

  12. [21]

    Semantic communications for wireless sensing: RIS-aided encoding and self-supervised decoding,

    H. Du, J. Wang, D. Niyato, J. Kang, Z. Xiong, J. Zhang, and X. Shen, “Semantic communications for wireless sensing: RIS-aided encoding and self-supervised decoding,” IEEE J. Sel. Areas Commun. , vol. 41, no. 8, pp. 2547–2562, 2023

    2023

  13. [22]

    Deep learnin g enabled semantic-secure communication with shuffling,

    F. Chen, L. Xiang, H. V . Cheng, and K. Shen, “Deep learnin g enabled semantic-secure communication with shuffling,” in IEEE Global Com- mun. Conf. (GLOBECOM) , 2023, pp. 6338–6343

    2023

  14. [23]

    Task-oriented multi-user semantic communications,

    H. Xie, Z. Qin, X. Tao, and K. B. Letaief, “Task-oriented multi-user semantic communications,” IEEE J. Sel. Areas Commun. , vol. 40, no. 9, pp. 2584–2597, 2022

    2022

  15. [24]

    C hannel- transferable semantic communications for multi-user OFDM -NOMA systems,

    L. Lin, W. Xu, F. Wang, Y . Zhang, W. Zhang, and P . Zhang, “C hannel- transferable semantic communications for multi-user OFDM -NOMA systems,” IEEE Wireless Commun. Lett. , 2023

    2023

  16. [25]

    Semantic feature division multiple access for mul ti-user digital interference networks,

    S. Ma, C. Zhang, B. Shen, Y . Wu, H. Li, S. Li, G. Shi, and N. A l- Dhahir, “Semantic feature division multiple access for mul ti-user digital interference networks,” IEEE Trans. Wireless Commun. , 2024

    2024

  17. [26]

    Flag vect or assisted multi-user semantic communications for downlink text tran smission,

    W. Huang, J. Wang, X. Chen, Q. Peng, and Y . Zhu, “Flag vect or assisted multi-user semantic communications for downlink text tran smission,” IEEE Commun. Lett. , 2024

    2024

  18. [27]

    Exploiting semantic communication fo r non- orthogonal multiple access,

    X. Mu and Y . Liu, “Exploiting semantic communication fo r non- orthogonal multiple access,” IEEE J. Sel. Areas Commun. , vol. 41, no. 8, pp. 2563–2576, 2023. 14

    2023

  19. [28]

    Fusion-base d multi-user semantic communications for wireless image transmission o ver degraded broadcast channels,

    T. Wu, Z. Chen, M. Tao, B. Xia, and W. Zhang, “Fusion-base d multi-user semantic communications for wireless image transmission o ver degraded broadcast channels,” in IEEE Global Commun. Conf. (GLOBECOM) , 2023, pp. 7623–7628

    2023

  20. [29]

    Deep joint source-channel coding for adaptive image transmissi on over MIMO channels,

    H. Wu, Y . Shao, C. Bian, K. Mikolajczyk, and D. G¨ und¨ uz, “Deep joint source-channel coding for adaptive image transmissi on over MIMO channels,” IEEE Trans. Wireless Commun. , 2024

    2024

  21. [30]

    Attention is all you need,

    A. V aswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jone s, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” in Neural Info. Process. Syst. (NeurIPS) , vol. 30, 2017

    2017

  22. [31]

    A sp atial- temporal transformer network for city-level cellular traf fic analysis and prediction,

    B. Gu, J. Zhan, S. Gong, W. Liu, Z. Su, and M. Guizani, “A sp atial- temporal transformer network for city-level cellular traf fic analysis and prediction,” IEEE Trans. Wireless Commun. , vol. 22, no. 12, pp. 9412– 9423, 2023

    2023

  23. [32]

    Fedfor mer: Frequency enhanced decomposed transformer for long-term s eries fore- casting,

    T. Zhou, Z. Ma, Q. Wen, X. Wang, L. Sun, and R. Jin, “Fedfor mer: Frequency enhanced decomposed transformer for long-term s eries fore- casting,” in Int’l Conf. Machine Learning (ICML) , 2022, pp. 27 268– 27 286

    2022

  24. [33]

    Transformer-aided wireless image transmission with chan nel feedback,

    H. Wu, Y . Shao, E. Ozfatura, K. Mikolajczyk, and D. G¨ und ¨ uz, “Transformer-aided wireless image transmission with chan nel feedback,” IEEE Trans. Wireless Commun. , 2024

    2024

  25. [34]

    Quan-transfo rmer based channel feedback for RIS-aided wireless communication sys tems,

    W. Xie, J. Zou, J. Xiao, M. Li, and X. Peng, “Quan-transfo rmer based channel feedback for RIS-aided wireless communication sys tems,” IEEE Commun. Lett. , vol. 26, no. 11, pp. 2631–2635, 2022

    2022

  26. [35]

    A robust deep learning enabled semantic communication syste m for text,

    X. Peng, Z. Qin, D. Huang, X. Tao, J. Lu, G. Liu, and C. Pan, “A robust deep learning enabled semantic communication syste m for text,” in IEEE Global Commun. Conf. (GLOBECOM) , 2022, pp. 2704–2709

    2022

  27. [36]

    Semantic communication wit h memory,

    H. Xie, Z. Qin, and G. Y . Li, “Semantic communication wit h memory,” IEEE J. Sel. Areas Commun. , vol. 41, no. 8, pp. 2658–2669, 2023

    2023

  28. [37]

    A lite distributed semantic communic ation system for internet of things,

    H. Xie and Z. Qin, “A lite distributed semantic communic ation system for internet of things,” IEEE J. Sel. Areas Commun. , vol. 39, no. 1, pp. 142–153, 2020

    2020

  29. [38]

    Semantic impo rtance- aware based for multi-user communication over MIMO fading c han- nels,

    H. Liang, Z. Bao, W. An, C. Dong, and X. Xu, “Semantic impo rtance- aware based for multi-user communication over MIMO fading c han- nels,” arXiv preprint arXiv:2312.16057 , 2023

    2023 arXiv

  30. [39]

    Sentence-BERT: Sentence e mbed- dings using Siamese BERT-networks,

    N. Reimers and I. Gurevych, “Sentence-BERT: Sentence e mbed- dings using Siamese BERT-networks,” 2019, [Online]. Avail able: https://arxiv.org/abs/1908.10084

    2019 arXiv

  31. [40]

    Semantic cosine simi- larity,

    F. Rahutomo, T. Kitasuka, M. Aritsugi et al. , “Semantic cosine simi- larity,” in Proc. 7th Int. Student Conf Adv. Sci. Technol. , vol. 4, no. 1. University of Seoul South Korea, 2012

    2012

  32. [41]

    BLEU: A m ethod for automatic evaluation of machine translation,

    K. Papineni, S. Roukos, T. Ward, and W.-J. Zhu, “BLEU: A m ethod for automatic evaluation of machine translation,” in Proc. Annual Meeting Assoc. Comput. Linguistics (ACL’02) , 2002, p. 311–318

    2002

  33. [42]

    Cider : Consensus- based image description evaluation,

    R. V edantam, C. Lawrence Zitnick, and D. Parikh, “Cider : Consensus- based image description evaluation,” in in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2015, pp. 4566–4575

    2015

  34. [43]

    GPT-4 technical report,

    J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F. L . Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkat et al. , “GPT-4 technical report,” arXiv preprint arXiv:2303.08774 , 2023

    2023 arXiv

  35. [44]

    Deep joint source chann el coding for wireless image transmission with OFDM,

    M. Y ang, C. Bian, and H.-S. Kim, “Deep joint source chann el coding for wireless image transmission with OFDM,” in Proc. IEEE Int. Conf. Commun. (ICC), , 2021, pp. 1–6

    2021

  36. [45]

    End-to-en d optimized image compression,

    J. Ball´ e, V . Laparra, and E. P . Simoncelli, “End-to-en d optimized image compression,” in 5th Proc. Int. Conf. Learn. Represent. (ICLR) , 2017

    2017

  37. [46]

    Adam: A method for stochastic opt imization,

    D. P . Kingma and J. Ba, “Adam: A method for stochastic opt imization,” 2014, [Online]. Available: https://arxiv.org/abs/1412. 6980

    2014

  38. [47]

    A la rge an- notated corpus for learning natural language inference,

    S. R. Bowman, G. Angeli, C. Potts, and C. D. Manning, “A la rge an- notated corpus for learning natural language inference,” 2 015, [Online]. Available: https://arxiv.org/abs/2201.01389

    arXiv

  39. [48]

    Europarl: A parallel corpus for statistical machine transla- tion,

    P . Koehn, “Europarl: A parallel corpus for statistical machine transla- tion,” in Proc. MT Summit , Sep, 2005, pp. 79–86

    2005

  40. [49]

    S eq2Seq-SC: End-to-end semantic communication systems with pre-train ed language model,

    J.-H. Lee, D.-H. Lee, E. Sheen, T. Choi, and J. Pujara, “S eq2Seq-SC: End-to-end semantic communication systems with pre-train ed language model,” in Proc. IEEE 57th Asilomar Conf. Signals, Syst. Comput. , 2023, pp. 260–264

    2023

  41. [50]

    Usin g Siamese BiLSTM models for identifying text semantic similarity,

    G. Fradelos, I. Perikos, and I. Hatzilygeroudis, “Usin g Siamese BiLSTM models for identifying text semantic similarity,” in IFIP AIAI WG 12.5 International W orkshops, 2023, pp. 381–392

    2023

  42. [51]

    Training for d iversity in image paragraph captioning,

    L. Melas-Kyriazi, A. M. Rush, and G. Han, “Training for d iversity in image paragraph captioning,” in Proc. Conf. Empir . Methods Natural Lang. Process. , 2018, pp. 757–761

    2018

  43. [52]

    A method for the construction of minimum -redundancy codes,

    D. A. Huffman, “A method for the construction of minimum -redundancy codes,” Proc. Inst. Radio Eng. , vol. 40, no. 9, pp. 1098–1101, Sep, 1952

    1952

  44. [53]

    An introduction to LDPC codes,

    W. E. Ryan et al. , “An introduction to LDPC codes,” CRC Handbook for Coding and Signal Processing for Recording Systems , vol. 5, no. 2, pp. 1–23, 2004

    2004

  45. [54]

    Design of polar codes in 5G new radio,

    V . Bioglio, C. Condo, and I. Land, “Design of polar codes in 5G new radio,” IEEE Commun. Surveys Tuts. , vol. 23, no. 1, pp. 29–40, 2020. Mingxiao Li (Graduate Student Member, IEEE) received the B.E. degree from Beijing University of Posts and Telecommunications, B eijing, Ch...

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.