pith. machine review for the scientific record. sign in

arxiv: 2601.10175 · v3 · submitted 2026-01-15 · 💻 cs.IT · math.IT

Recognition: no theorem link

A Low-Complexity Framework for Multi-access Coded Caching Systems with Arbitrary User-cache Access Topology

Ting Yang , Kai Wan , Minquan Cheng , Xinping Yi , Robert Caiming Qiu , Giuseppe Caire
Authors on Pith no claims yet

Pith reviewed 2026-05-16 14:37 UTC · model grok-4.3

classification 💻 cs.IT math.IT
keywords multi-access coded cachingarbitrary access topologyconflict graphgraph coloringgraph neural networksindex coding boundtransmission loadlow-complexity delivery
0
0 comments X

The pith

Graph neural networks color conflict graphs to deliver multi-access coded caching at loads close to the index-coding bound with low complexity.

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

The paper models the delivery phase of multi-access coded caching with arbitrary user-cache access patterns as a graph coloring task on a conflict graph whose chromatic number sets the transmission load. A graph neural network is trained to produce efficient colorings that yield coded multicast messages, and this model is shown to generalize across different numbers of users and unseen access topologies. The resulting loads nearly match those of the exact DSatur algorithm and an extended index-coding lower bound while requiring far less computation. A separate low-complexity greedy algorithm is also derived that approximates the same bound.

Core claim

By representing packet delivery conflicts in an arbitrary MACC system as edges in a conflict graph and using a GNN to assign colors, the framework constructs a set of coded transmissions whose total length is close to the information-theoretic minimum for any given user-cache access topology and request pattern.

What carries the argument

The conflict graph whose vertices are requested packets and whose edges encode decoding conflicts among users; its minimum coloring directly determines the minimal number of multicast transmissions needed.

If this is right

  • The GNN method scales to system sizes where exact coloring algorithms become intractable.
  • Coded caching remains feasible for arbitrary rather than only combinatorially structured access patterns.
  • The greedy approximation supplies a simple baseline that already lies close to the information-theoretic limit.
  • Delivery schemes can be computed once per topology class and reused without redesign.

Where Pith is reading between the lines

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

  • Learned graph-coloring heuristics may replace combinatorial solvers in other index-coding or network-coding settings that admit a conflict-graph formulation.
  • If the generalization holds, the same trained model could support online topology changes without full retraining.
  • The reduction to graph coloring suggests that progress on scalable graph algorithms directly improves practical coded caching performance.

Load-bearing premise

A single GNN trained on a finite collection of topologies and user counts will continue to produce near-optimal colorings on completely new arbitrary access topologies without retraining.

What would settle it

A test set of previously unseen access topologies and user counts on which the GNN's achieved transmission load exceeds the extended index-coding bound by a margin larger than the gap observed with DSatur.

Figures

Figures reproduced from arXiv: 2601.10175 by Giuseppe Caire, Kai Wan, Minquan Cheng, Robert Caiming Qiu, Ting Yang, Xinping Yi.

Figure 1
Figure 1. Figure 1: MACC system model with arbitrary user-cache access topology. [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: User-cache access topology with 5 users and 4 cache nodes. given by A = {Ak | k ∈ [5]}, where A1 = {1, 2}, A2 = {1, 3}, A3 = {4}, A4 = {2}, A5 = {3}. (3) Each cache node stores the following subsets of packets according to the array P in [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Transformation from the MN PDA P to an MACC scheme via arrays C, U, and Q. access topology. To the best of our knowledge, a general coded caching scheme that can accommodate MACC systems with arbitrary access topology is still lacking. This observation motivates us to investigate a more general and flexible design framework. In the following, we show that the user-delivery array Q, constructed based on the… view at source ↗
Figure 4
Figure 4. Figure 4: Transformation from the user-retrieve array [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Step 1, detailed in Section II-B, transforms the MACC-PDA problem into an equivalent [PITH_FULL_IMAGE:figures/full_fig_p021_5.png] view at source ↗
Figure 5
Figure 5. Figure 5: AI-aided framework for the delivery phase in MACC systems with arbitrary user–cache [PITH_FULL_IMAGE:figures/full_fig_p022_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Overall architecture of the proposed GNN model for graph coloring in MACC systems. [PITH_FULL_IMAGE:figures/full_fig_p024_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The proposed postprocessing conflict repair procedure. [PITH_FULL_IMAGE:figures/full_fig_p030_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The procedure of greedy converse. Consequently, Line 12 selects a column with the largest intersection cardinality. Among columns 3, 4, 5, which all attain the maximum value, we arbitrarily select k ′ = 3. Follow￾ing Lines 13–15, the accumulated set is updated to S = I3, the cumulative sum becomes S = |S| = 3, and column 3 is removed from Kremain. This greedy selection procedure is repeated until all colum… view at source ↗
Figure 9
Figure 9. Figure 9: Ratio between the number of colors produced by DSatur in Algorithm 1 and the IC [PITH_FULL_IMAGE:figures/full_fig_p038_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Performance comparison of the proposed GNN-based approach and the GIN-based [PITH_FULL_IMAGE:figures/full_fig_p039_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Performance of the proposed GNN-based approach across varying numbers of user [PITH_FULL_IMAGE:figures/full_fig_p040_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Performance comparison between the proposed greedy converse and the IC converse [PITH_FULL_IMAGE:figures/full_fig_p041_12.png] view at source ↗
read the original abstract

This paper studies the multi-access coded caching (MACC) problem with arbitrary user-cache access topology, which extends existing MACC models that rely on highly structured and combinatorially designed topologies. We consider a MACC system consisting of a single server, $\Lambda$ cache-nodes, and $K$ user-nodes. The server stores $N$ equal-size files, each cache-node has a storage capacity of $M$ files, and each user-node $k\in[K]$ can access an arbitrary subset of cache-nodes $\mathcal{A}_k\subseteq[\Lambda]$ and retrieve the cached content stored in cache-nodes $\mathcal{A}_k$. The objective is to design a universal framework for the MACC delivery problem. Decoding conflicts among the requested packets are captured by a conflict graph, and the design of the delivery is reduced to a graph coloring problem, where achieving a lower transmission load corresponds to coloring the graph using fewer colors. Under this formulation, the classical DSatur algorithm achieves a transmission load close to the index-coding (IC) converse bound, thereby providing a practical benchmark. However, its computational complexity becomes prohibitive for large-scale graphs. To overcome this limitation, we develop a learning-driven approach using graph neural networks (GNNs) that efficiently constructs coded multicast transmissions with performance close to the theoretical bounds and generalizes across different user-cache access topologies and numbers of users. In addition, we extend the IC converse bound to MACC systems with arbitrary access topology and propose a low-complexity greedy approximation that closely matches the IC converse bound. Numerical results demonstrate that the proposed approach achieves performance close to the DSatur algorithm and the IC converse bound, while significantly reducing computational complexity, making it well-suited for large-scale MACC systems.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The paper studies multi-access coded caching (MACC) with arbitrary user-cache access topologies A_k. It models the delivery phase as a graph-coloring problem on the induced conflict graph, replaces the high-complexity DSatur heuristic with a GNN-based solver, extends the index-coding converse bound to this setting, and introduces a low-complexity greedy approximation to the bound. Numerical experiments are claimed to show that the GNN produces transmission loads close to both DSatur and the extended IC bound while scaling to large K and Lambda.

Significance. If the GNN generalizes reliably, the work supplies a practical, low-complexity algorithmic framework for MACC systems whose access patterns are not combinatorially structured, together with a matching converse and a simple greedy benchmark. The reduction to conflict-graph coloring and the explicit extension of the IC bound are technically clean contributions that could be reused beyond the GNN component.

major comments (2)
  1. [Numerical Results] Numerical Results section: the headline claim that the GNN 'achieves performance close to the DSatur algorithm and the IC converse bound' is stated without any tabulated transmission-load values, ratios, standard deviations, or explicit description of the test-set generation process (distribution over Lambda, K, and A_k). Because the central performance assertion rests on these numbers, their absence prevents verification of the generalization claim.
  2. [GNN Approach] GNN training and generalization paragraph: the manuscript asserts that a single trained model works for 'different user-cache access topologies and numbers of users,' yet provides no details on the training distribution, whether test instances are drawn from the same generative process, or any out-of-distribution experiments with pathological A_k or K larger than the training range. This is load-bearing for the 'arbitrary topology' claim.
minor comments (2)
  1. [Abstract] The abstract and introduction use 'close to' without defining a quantitative tolerance; a precise statement (e.g., within X % of DSatur load) would improve clarity.
  2. [IC Bound Extension] Notation for the extended IC bound is introduced without an explicit equation number in the main text; cross-referencing the bound expression would help readers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which highlight important aspects of clarity and verifiability in our numerical evaluation and GNN generalization claims. We address each point below and will revise the manuscript to incorporate the requested details.

read point-by-point responses

  1. Referee: [Numerical Results] Numerical Results section: the headline claim that the GNN 'achieves performance close to the DSatur algorithm and the IC converse bound' is stated without any tabulated transmission-load values, ratios, standard deviations, or explicit description of the test-set generation process (distribution over Lambda, K, and A_k). Because the central performance assertion rests on these numbers, their absence prevents verification of the generalization claim.

    Authors: We agree that the current presentation of the numerical results lacks sufficient quantitative detail for independent verification. In the revised manuscript we will add explicit tables reporting average transmission loads (with ratios to both DSatur and the extended IC bound), standard deviations over repeated trials, and a precise description of the test-set generation process, including the ranges and sampling distributions used for Lambda, K, and the random access topologies A_k. revision: yes

  2. Referee: [GNN Approach] GNN training and generalization paragraph: the manuscript asserts that a single trained model works for 'different user-cache access topologies and numbers of users,' yet provides no details on the training distribution, whether test instances are drawn from the same generative process, or any out-of-distribution experiments with pathological A_k or K larger than the training range. This is load-bearing for the 'arbitrary topology' claim.

    Authors: We acknowledge the need for greater transparency on the training regime to substantiate the generalization statements. The revised version will specify the exact training distribution over topologies and system sizes, confirm that the reported test instances are drawn from the same generative process, and add out-of-distribution experiments (including larger K and atypical or pathological A_k) to quantify the model's robustness beyond the training range. revision: yes

Circularity Check

0 steps flagged

No circularity; standard graph reduction plus empirical GNN heuristic with separate training/test sets

full rationale

The derivation reduces MACC delivery to coloring the conflict graph induced by arbitrary A_k access sets (standard modeling step, not self-referential). The IC converse bound is extended mathematically to the arbitrary-topology case and used as an external benchmark rather than fitted from the same numerical results. DSatur serves as an independent algorithmic reference. The GNN is trained on finite sampled topologies and evaluated on held-out instances; reported closeness is an empirical observation, not a quantity defined by construction from the training data or from the bound itself. No equation, parameter fit, or self-citation chain collapses the claimed transmission load back onto the inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The framework relies on standard graph theory (conflict graphs are colorable) and information-theoretic index-coding bounds; no new physical constants or ad-hoc fitted scalars are introduced in the abstract description.

axioms (2)
  • domain assumption The delivery problem can be exactly modeled as a conflict graph whose chromatic number equals the minimum number of multicast transmissions.
    Invoked when the authors reduce the MACC delivery to graph coloring.
  • domain assumption The index-coding converse bound remains valid after extension to arbitrary access topologies.
    Stated as an extension of the classical IC bound.

pith-pipeline@v0.9.0 · 5637 in / 1382 out tokens · 52100 ms · 2026-05-16T14:37:23.978942+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

54 extracted references · 54 canonical work pages · 1 internal anchor

  1. [1]

    The role of caching in future communication systems and networks,

    G. S. Paschos, G. Iosifidis, M. Tao, D. Towsley, and G. Caire, “The role of caching in future communication systems and networks,”IEEE Journal on Selected Areas in Communications, vol. 36, no. 6, pp. 1111–1125, 2018

    work page 2018

  2. [2]

    Fundamental limits of caching,

    M. A. Maddah-Ali and U. Niesen, “Fundamental limits of caching,”IEEE Transactions on information theory, vol. 60, no. 5, pp. 2856–2867, 2014

    work page 2014

  3. [3]

    An index coding approach to caching with uncoded cache placement,

    K. Wan, D. Tuninetti, and P. Piantanida, “An index coding approach to caching with uncoded cache placement,”IEEE Transactions on Information Theory, vol. 66, no. 3, pp. 1318–1332, 2020

    work page 2020

  4. [4]

    On the placement delivery array design for centralized coded caching scheme,

    Q. Yan, M. Cheng, X. Tang, and Q. Chen, “On the placement delivery array design for centralized coded caching scheme,” IEEE Transactions on Information Theory, vol. 63, no. 9, pp. 5821–5833, 2017

    work page 2017

  5. [5]

    A framework of constructing placement delivery arrays for centralized coded caching,

    M. Cheng, J. Wang, X. Zhong, and Q. Wang, “A framework of constructing placement delivery arrays for centralized coded caching,”IEEE Transactions on Information Theory, vol. 67, no. 11, pp. 7121–7131, 2021. 43

    work page 2021

  6. [6]

    Placement delivery array construction via cartesian product for coded caching,

    J. Wang, M. Cheng, K. Wan, and G. Caire, “Placement delivery array construction via cartesian product for coded caching,” IEEE Transactions on Information Theory, vol. 69, no. 12, pp. 7602–7626, 2023

    work page 2023

  7. [7]

    A framework of constructing pda via union of cache configurations from cartesian product,

    ——, “A framework of constructing pda via union of cache configurations from cartesian product,” in2025 IEEE International Symposium on Information Theory (ISIT). IEEE, 2025, pp. 1–6

    work page 2025

  8. [8]

    A novel construction of coded caching schemes with polynomial subpacketizations via projective geometry,

    H. Wei, M. Cheng, and K. Leung, “A novel construction of coded caching schemes with polynomial subpacketizations via projective geometry,” in2024 IEEE Information Theory Workshop (ITW). IEEE, 2024, pp. 496–501

    work page 2024

  9. [9]

    Improved constructions of coded caching schemes for combination networks,

    M. Cheng, Y . Li, X. Zhong, and R. Wei, “Improved constructions of coded caching schemes for combination networks,” IEEE Transactions on Communications, vol. 68, no. 10, pp. 5965–5975, 2020

    work page 2020

  10. [10]

    Fundamental limits of caching in wireless d2d networks,

    M. Ji, G. Caire, and A. F. Molisch, “Fundamental limits of caching in wireless d2d networks,”IEEE Transactions on Information Theory, vol. 62, no. 2, pp. 849–869, 2015

    work page 2015

  11. [11]

    Combinatorial designs for coded caching on hierarchical networks,

    Y . Kong, Y . Wu, and M. Cheng, “Combinatorial designs for coded caching on hierarchical networks,” in2023 IEEE Wireless Communications and Networking Conference (WCNC). IEEE, 2023, pp. 1–6

    work page 2023

  12. [12]

    Multiple-antenna placement delivery array for cache-aided miso systems,

    T. Yang, K. Wan, M. Cheng, R. C. Qiu, and G. Caire, “Multiple-antenna placement delivery array for cache-aided miso systems,”IEEE Transactions on Information Theory, 2023

    work page 2023

  13. [13]

    Adding transmitters dramatically boosts coded-caching gains for finite file sizes,

    E. Lampiris and P. Elia, “Adding transmitters dramatically boosts coded-caching gains for finite file sizes,”IEEE Journal on Selected Areas in Communications, vol. 36, no. 6, pp. 1176–1188, 2018

    work page 2018

  14. [14]

    Fundamental limits of cache-aided interference management,

    N. Naderializadeh, M. A. Maddah-Ali, and A. S. Avestimehr, “Fundamental limits of cache-aided interference management,” IEEE Transactions on Information Theory, vol. 63, no. 5, pp. 3092–3107, 2017

    work page 2017

  15. [15]

    Extended placement delivery arrays for multi-antenna coded caching scheme,

    K. K. Namboodiri, E. Peter, and B. S. Rajan, “Extended placement delivery arrays for multi-antenna coded caching scheme,”IEEE Transactions on Communications, 2023

    work page 2023

  16. [16]

    Low-complexity high-performance cyclic caching for large miso systems,

    M. Salehi, E. Parrinello, S. P. Shariatpanahi, P. Elia, and A. Tölli, “Low-complexity high-performance cyclic caching for large miso systems,”IEEE Transactions on Wireless Communications, vol. 21, no. 5, pp. 3263–3278, 2021

    work page 2021

  17. [17]

    Reflecting intelligent surfaces-assisted multiple-antenna coded caching,

    X. Niu, M. Cheng, K. Wan, R. C. Qiu, and G. Caire, “Reflecting intelligent surfaces-assisted multiple-antenna coded caching,” in2024 IEEE Information Theory Workshop (ITW). IEEE, 2024, pp. 490–495

    work page 2024

  18. [18]

    A novel transformation approach of shared-link coded caching schemes for multiaccess networks,

    M. Cheng, K. Wan, D. Liang, M. Zhang, and G. Caire, “A novel transformation approach of shared-link coded caching schemes for multiaccess networks,”IEEE Transactions on Communications, vol. 69, no. 11, pp. 7376–7389, 2021

    work page 2021

  19. [19]

    Multi-access coded caching scheme with linear sub-packetization using pdas,

    S. Sasi and B. S. Rajan, “Multi-access coded caching scheme with linear sub-packetization using pdas,”IEEE Transactions on Communications, vol. 69, no. 12, pp. 7974–7985, 2021

    work page 2021

  20. [20]

    Multi-access coded caching with optimal rate and linear subpacketization under pda and consecutive cyclic placement,

    J. Wang, M. Cheng, Y . Wu, and X. Li, “Multi-access coded caching with optimal rate and linear subpacketization under pda and consecutive cyclic placement,”IEEE Transactions on Communications, vol. 71, no. 6, pp. 3178–3190, 2023

    work page 2023

  21. [21]

    Coded caching for two-dimensional multi-access networks,

    M. Zhang, K. Wan, M. Cheng, and G. Caire, “Coded caching for two-dimensional multi-access networks,” in2022 IEEE International Symposium on Information Theory (ISIT). IEEE, 2022, pp. 1707–1712

    work page 2022

  22. [22]

    Coded caching for multi-level popularity and access,

    J. Hachem, N. Karamchandani, and S. N. Diggavi, “Coded caching for multi-level popularity and access,”IEEE Transactions on Information Theory, vol. 63, no. 5, pp. 3108–3141, 2017

    work page 2017

  23. [23]

    Structured index coding problem and multi-access coded caching,

    K. S. Reddy and N. Karamchandani, “Structured index coding problem and multi-access coded caching,”IEEE Journal on Selected Areas in Information Theory, vol. 2, no. 4, pp. 1266–1281, 2021

    work page 2021

  24. [24]

    Fundamental limits of combinatorial multi-access caching,

    F. Brunero and P. Elia, “Fundamental limits of combinatorial multi-access caching,”IEEE Transactions on Information Theory, vol. 69, no. 2, pp. 1037–1056, 2022

    work page 2022

  25. [25]

    Combinatorial multi-access coded caching: Improved rate-memory trade-off with coded placement,

    K. K. Namboodiri and B. S. Rajan, “Combinatorial multi-access coded caching: Improved rate-memory trade-off with coded placement,”IEEE Transactions on Information Theory, vol. 70, no. 3, pp. 1787–1805, 2024

    work page 2024

  26. [26]

    Multi-access coded caching schemes from cross resolvable designs,

    D. Katyal, P. N. Muralidhar, and B. S. Rajan, “Multi-access coded caching schemes from cross resolvable designs,”IEEE Transactions on Communications, vol. 69, no. 5, pp. 2997–3010, 2021. 44

    work page 2021

  27. [27]

    Multi-access coded caching schemes from maximal cross resolvable designs,

    N. Das and B. S. Rajan, “Multi-access coded caching schemes from maximal cross resolvable designs,” in2022 IEEE International Symposium on Information Theory (ISIT). IEEE, 2022, pp. 1719–1724

    work page 2022

  28. [28]

    Multi-access coded caching from a new class of cross resolvable designs,

    P. N. Muralidhar and B. S. Rajan, “Multi-access coded caching from a new class of cross resolvable designs,” in2021 IEEE International Symposium on Information Theory (ISIT). IEEE, 2021, pp. 855–860

    work page 2021

  29. [29]

    Coded caching schemes for multiaccess topologies via combinatorial design,

    M. Cheng, K. Wan, P. Elia, and G. Caire, “Coded caching schemes for multiaccess topologies via combinatorial design,” IEEE Transactions on Information Theory, 2025

    work page 2025

  30. [30]

    Deep learning for wireless coded caching with unknown and time-variant content popularity,

    Z. Zhang and M. Tao, “Deep learning for wireless coded caching with unknown and time-variant content popularity,”IEEE Transactions on Wireless Communications, vol. 20, no. 2, pp. 1152–1163, 2020

    work page 2020

  31. [31]

    Deep reinforcement learning based coded caching scheme in fog radio access networks,

    Y . Zhou, M. Peng, S. Yan, and Y . Sun, “Deep reinforcement learning based coded caching scheme in fog radio access networks,” in2018 IEEE/CIC International Conference on Communications in China (ICCC Workshops). IEEE, 2018, pp. 309–313

    work page 2018

  32. [32]

    Learning distributed caching strategies in small cell networks,

    A. Sengupta, S. Amuru, R. Tandon, R. M. Buehrer, and T. C. Clancy, “Learning distributed caching strategies in small cell networks,” in2014 11th International Symposium on Wireless Communications Systems (ISWCS). IEEE, 2014, pp. 917–921

    work page 2014

  33. [33]

    Multi-agent reinforcement learning for cooperative coded caching via homotopy optimization,

    X. Wu, J. Li, M. Xiao, P. Ching, and H. V . Poor, “Multi-agent reinforcement learning for cooperative coded caching via homotopy optimization,”IEEE Transactions on Wireless Communications, vol. 20, no. 8, pp. 5258–5272, 2021

    work page 2021

  34. [34]

    Reinforcement learning based cooperative coded caching under dynamic popularities in ultra-dense networks,

    S. Gao, P. Dong, Z. Pan, and G. Y . Li, “Reinforcement learning based cooperative coded caching under dynamic popularities in ultra-dense networks,”IEEE Transactions on Vehicular Technology, vol. 69, no. 5, pp. 5442–5456, 2020

    work page 2020

  35. [35]

    Learning-based content caching and sharing for wireless networks,

    J. Song, M. Sheng, T. Q. Quek, C. Xu, and X. Wang, “Learning-based content caching and sharing for wireless networks,” IEEE Transactions on Communications, vol. 65, no. 10, pp. 4309–4324, 2017

    work page 2017

  36. [36]

    Double coded caching in ultra dense networks: Caching and multicast scheduling via deep reinforcement learning,

    Z. Zhang, H. Chen, M. Hua, C. Li, Y . Huang, and L. Yang, “Double coded caching in ultra dense networks: Caching and multicast scheduling via deep reinforcement learning,”IEEE Transactions on Communications, vol. 68, no. 2, pp. 1071–1086, 2019

    work page 2019

  37. [37]

    Learning to code: Coded caching via deep reinforcement learning,

    N. Naderializadeh and S. M. Asghari, “Learning to code: Coded caching via deep reinforcement learning,” in2019 53rd Asilomar Conference on Signals, Systems, and Computers. IEEE, 2019, pp. 1774–1778

    work page 2019

  38. [38]

    Revisiting topological interference management: A learning-to-code on graphs perspective,

    Z. Shan, X. Yi, H. Yu, C.-S. Liao, and S. Jin, “Revisiting topological interference management: A learning-to-code on graphs perspective,”arXiv preprint arXiv:2502.09344, 2025

    work page arXiv 2025

  39. [39]

    J. A. Bondy, U. S. R. Murtyet al.,Graph theory with applications. Macmillan London, 1976, vol. 290

    work page 1976

  40. [40]

    Computers and intractability: a guide to the theory of np-completeness (michael r. garey and david s. johnson),

    J. Hartmanis, “Computers and intractability: a guide to the theory of np-completeness (michael r. garey and david s. johnson),”Siam Review, vol. 24, no. 1, p. 90, 1982

    work page 1982

  41. [41]

    Maddah-ali-niesen scheme for multi-access coded caching,

    P. N. Muralidhar, D. Katyal, and B. S. Rajan, “Maddah-ali-niesen scheme for multi-access coded caching,” in2021 IEEE Information Theory Workshop (ITW). IEEE, 2021, pp. 1–6

    work page 2021

  42. [42]

    Rate-memory trade-off for multi-access coded caching with uncoded placement,

    K. S. Reddy and N. Karamchandani, “Rate-memory trade-off for multi-access coded caching with uncoded placement,” IEEE Transactions on Communications, vol. 68, no. 6, pp. 3261–3274, 2020

    work page 2020

  43. [43]

    New methods to color the vertices of a graph,

    D. Brélaz, “New methods to color the vertices of a graph,”Communications of the ACM, vol. 22, no. 4, pp. 251–256, 1979

    work page 1979

  44. [44]

    A deep learning guided memetic framework for graph coloring problems,

    O. Goudet, C. Grelier, and J.-K. Hao, “A deep learning guided memetic framework for graph coloring problems,” Knowledge-Based Systems, vol. 258, p. 109986, 2022

    work page 2022

  45. [45]

    Rethinking graph neural networks for graph coloring,

    W. Li, R. Li, Y . Ma, S. O. Chan, and B. Yu, “Rethinking graph neural networks for graph coloring,” 2020

    work page 2020

  46. [46]

    Graph coloring with physics-inspired graph neural networks,

    M. J. Schuetz, J. K. Brubaker, Z. Zhu, and H. G. Katzgraber, “Graph coloring with physics-inspired graph neural networks,” Physical Review Research, vol. 4, no. 4, p. 043131, 2022. 45

    work page 2022

  47. [47]

    Using graph isomorphism network to solve graph coloring problem,

    Y . Zhang, K. Zhang, and N. Wu, “Using graph isomorphism network to solve graph coloring problem,” in2024 4th International Conference on Consumer Electronics and Computer Engineering (ICCECE). IEEE, 2024, pp. 209–215

    work page 2024

  48. [48]

    The potts model,

    F.-Y . Wu, “The potts model,”Reviews of modern physics, vol. 54, no. 1, p. 235, 1982

    work page 1982

  49. [49]

    Semi-Supervised Classification with Graph Convolutional Networks

    T. Kipf, “Semi-supervised classification with graph convolutional networks,”arXiv preprint arXiv:1609.02907, 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  50. [50]

    Using tabu search techniques for graph coloring,

    A. Hertz and D. de Werra, “Using tabu search techniques for graph coloring,”Computing, vol. 39, no. 4, pp. 345–351, 1987

    work page 1987

  51. [51]

    Rethinking graph neural networks for the graph coloring problem,

    W. Li, R. Li, Y . Ma, S. O. Chan, D. Pan, and B. Yu, “Rethinking graph neural networks for the graph coloring problem,” arXiv preprint arXiv:2208.06975, 2022

    work page arXiv 2022

  52. [52]

    Graph coloring algorithm based on minimal cost graph neural network,

    M. Gao and J. Hu, “Graph coloring algorithm based on minimal cost graph neural network,”IEEE Access, 2024

    work page 2024

  53. [53]

    Solving graph coloring problem via graph neural network (gnn),

    A. Z. Ijaz, R. H. Ali, N. Ali, T. Laique, and T. A. Khan, “Solving graph coloring problem via graph neural network (gnn),” in2022 17th International Conference on Emerging Technologies (ICET). IEEE, 2022, pp. 178–183

    work page 2022

  54. [54]

    On the capacity region for index coding,

    F. Arbabjolfaei, B. Bandemer, Y .-H. Kim, E. ¸ Sa¸ so˘glu, and L. Wang, “On the capacity region for index coding,” in2013 IEEE International Symposium on Information Theory. IEEE, 2013, pp. 962–966

    work page 2013