Pith. sign in

REVIEW 4 minor 36 references

Square-Root Law for Covert Communication with Warden-Favorable Side Information

T0 review · 0 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read The square-root-law constant for covert communication is set by the variance of the warden's residual after cancellation, not by his receiver noise alone.

desk verdict A clean, self-contained refinement of the SRL constant: when Willie cancels a public component and the residual is Gaussian with known variance, the first-order covert throughput is governed by the post-cancellation variance σ_0², not the receiver noise alone. read the letter →

arxiv 2607.14013 v1 pith:GFFX4AQA submitted 2026-07-15 cs.IT math.IT

Hossein Ahmadi , Christian Deppe , Boulat A. Bash , Eduard A. Jorswieck This is my paper
classification cs.ITmath.IT
keywords covertcommunicationsquare-rootlawrelativeentropylowprobabilityofdetectionsideinformationresidualfloorGaussianinnovationpowerallocation
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 studies covert communication when the warden, Willie, has access to all physically obtainable side information—protocol details, timing, pilots, channel estimates—and cancels the aggregate public component of his received signal before testing for covert transmission. The paper's central claim is that, under this warden-favorable residualization, the maximal covert payload over n channel uses is asymptotically R_C^*√n bits with R_C^* = σ_0²/(σ_B² ln 2)√δ, where σ_0² = σ_W² + σ_e² is the variance of Willie's post-cancellation residual under no covert transmission, i.e., his receiver noise plus an irreducible cancellation error. The paper proves achievability with low-power Gaussian signaling and a matching converse showing that no input distribution—sparse, on-off, or otherwise—can improve the first-order constant within the assumed Gaussian innovation model. The result matters because it shows the square-root law's constant is governed by the variance at Willie's actual detector input, not by his receiver noise alone, and it yields a concrete power-allocation rule, the σ⁴ scheduler, when the residual variance varies over time.

What carries the argument

Willie's innovation residual R^n = W^n − \hat{U}_W^n is the central statistic; covertness is enforced by the conditioned relative-entropy constraint D(P_{R^n|H_1,S=s} ∥ P_{R^n|H_0,S=s}) ≤ δ. The argument is carried by Assumption A3, the Gaussian innovation model: given S=s, the null residual is N(0,σ_0²I_n) and the covert residual is X_c^n plus that same independent noise. This reduces the relative-entropy budget to ½Σ(ρ_t − ln(1+ρ_t)) ≤ δ, whose quadratic bounds yield the ℓ² power bound ΣP_t² ≤ 4σ_0⁴δ(1+o(1)); with the AWGN mutual-information bound at Bob and Cauchy–Schwarz, this gives the exact constant. In the heterogeneous case the same budget becomes the ellipsoidal constraint Σ P_t²/(4

What would settle it

Numerically compute D(P_{X+N}∥N(0,σ_0²I_n)) for a sparse/on-off X with coordinate variances P_t: if any such X gives a relative entropy strictly below the Gaussian bound ½Σ(ρ_t−ln(1+ρ_t)), Lemma 4 is refuted and the converse fails; a hardware measurement of the residual distribution that shows non-Gaussianity or unknown variance would likewise place the system outside the theorem's scope.

Watch

Extended reading notes

Core claim

Theorem 5 establishes the exact first-order scaling: B_n^* = R_C^*√n(1+o(1)), with R_C^* = σ_0²/(σ_B² ln 2)√δ. Achievability uses i.i.d. Gaussian covert symbols at power P_n = 2σ_0²√(δ/n), which meet the relative-entropy budget with the equalizer schedule. The converse rests on Lemma 4, which states that for a fixed coordinate-variance profile, the independent Gaussian variance-increase law minimizes relative entropy to the Gaussian null law; hence sparse or on-off signaling cannot do better. Consequently, the residual floor σ_e² multiplies the baseline SRL constant by (1+σ_e²/σ_W²).

Load-bearing premise

The load-bearing premise is Assumption (A3): conditioned on the warden's side information, the post-cancellation residual under no covert transmission is zero-mean Gaussian with a known variance profile; if cancellation leaves a non-Gaussian law or fixed non-vanishing variance uncertainty, the square-root-law constant and the entire achievability–converse argument no longer follow.

Editorial extensions

If this is right

  • In the stationary residual case, the exact first-order throughput is B_n^* = (σ_0²/(σ_B² ln 2))√(nδ)(1+o(1)), achieved by i.i.d. Gaussian symbols at the equalizer power.
  • No input distribution—sparse, on-off, or otherwise—can improve the first-order constant under the additive Gaussian innovation model, because Gaussian residuals are least detectable for fixed coordinate variances (Lemma 4).
  • When conditioned residual variances vary across channel uses, the first-order optimal allocation is P_t ∝ σ_{0,t}^4, yielding a total covert power that scales as 2√δ√(Σ σ_{0,t}^4).
  • A conservative design that enforces covertness on the full pre-cancellation observation recovers only the baseline constant σ_W²/(σ_B² ln 2)√δ, confirming that the residual-floor boost requires the warden's cancellation step to be effective.

Reading between the lines

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

  • The gain factor (1+σ_e²/σ_W²) suggests a design lever the paper does not develop: if a covert system can deliberately make the public component harder for Willie to cancel, the residual floor σ_e² grows and the covert rate constant rises, at the price of a more burdensome public layer.
  • The extremality lemma is stated for a diagonal Gaussian null; a natural extension is colored noise with a known covariance Σ_0, where the σ⁴ scheduler would plausibly become a waterfilling over the eigenmodes of Σ_0.
  • The paper's regime boundary identifies a testable prediction: in systems where the post-cancellation residual is heavy-tailed or its variance is uncertain, the square-root-law constant may fail; measuring the residual distribution in a prototype cancellation system would indicate whether the Gaussian innovation model or a noise-uncertainty/linear-law regime applies.
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.

Referee Report

0 major / 4 minor

Summary. The paper studies covert communication over a scalar Gaussian overlay channel in which Willie is given all physically obtainable side information S and performs block-level cancellation of the public/overt component before testing. Covertness is imposed on the post-cancellation residual through a conditioned relative-entropy constraint D(P_{R^n|H_1,S=s} || P_{R^n|H_0,S=s}) ≤ δ for all admissible s. Under assumptions (A1)–(A4), the key one being (A3) that the post-cancellation null residual is zero-mean Gaussian with known conditioned variance σ_0^2 = σ_W^2 + σ_e^2, the paper proves Theorem 5: the maximal covert payload is B_n^* = R_C^* √n (1+o(1)) with R_C^* = σ_0^2/(σ_B^2 ln 2) √δ bits. Achievability uses i.i.d. Gaussian signaling with P_n = 2σ_0^2√(δ/n); the converse is based on Lemma 4 (Gaussian variance increase is least detectable), an ℓ_2 power bound, Cauchy–Schwarz, and the AWGN mutual-information bound at Bob. The paper also derives a heterogeneous σ^4-type optimal scheduler and a conservative pre-residualization design that recovers the baseline constant σ_W^2/(σ_B^2 ln 2)√δ.

Significance. If the result holds, it gives a clean first-order generalization of the classical AWGN square-root law to warden-favorable settings with side information and cancellation: the SRL constant is governed by the variance of Willie's actual detector input after cancellation, not by receiver noise alone. The derivation is self-contained and sound: the relative-entropy identity for variance increases, the convexity-based equalizer, the Gaussian-extremality lemma, and the subsequent ℓ_2→AWGN bound are all correct. A particular strength is that the theorem is explicitly conditional: Assumption (A3) is advertised as an achievability/regime model, not as a universal property of all cancellers, and the paper repeatedly states the boundary at which it fails (Sections II-A, IV-C, condition T2). The main caveat is external validity—real cancellation residuals may be non-Gaussian or may depend on the covert signal—but this is a stated limitation rather than an internal inconsistency. The paper also correctly identifies the conservative pre-residualization design as recovering the baseline constant, which sharpens the interpretation of the boost σ_0^2/σ_W^2. I regard the central claim as internally

minor comments (4)
  1. [Section III-F, Eq. (29)–(32)] The heterogeneous-case converse is stated in a single sentence ('the same Cauchy–Schwarz argument gives the matching first-order converse'). Since Lemma 4 is proved only for identical coordinate variances, it would improve self-containedness to state and prove the natural generalization D(P_R ∥ N(0, diag(σ_{0,t}^2))) ≥ (1/2)Σ(ρ_t − ln(1+ρ_t)) or to give the few lines of the heterogeneous ℓ_2-to-AWGN argument.
  2. [Section II-A, Assumption (A3)] The condition that the same innovation vector N^n(s) is independent of X_c^n under H_1 is the load-bearing idealization, and the paper flags it well. Because the title emphasizes 'warden-favorable side information,' a reader might initially expect the paper to optimize over all cancellers; adding one sentence in Section I explicitly stating that no claim is made for cancellers that violate (A3) (e.g., full-block linear MMSE cancellation generally induces an X_c-dependent residual) would further reduce the risk of over-interpretation.
  3. [Section II-B] The dual notation P_{R^n_i|S=s} and P_{R^n|H_i,S=s} is used interchangeably. It is understandable but slightly redundant; using one convention throughout would improve readability.
  4. [Figure 3] The horizontal axis for Figure 3 is labeled 'Covertness budget δ' but the tick labels appear to be powers of 10; please clarify whether the axis is logarithmic, and whether the exact equalizer curve uses the exact ρ_n or the closed-form allocation in the normalization.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: Theorem 5 is a self-contained conditional characterization under the paper's explicitly stated Gaussian-innovation assumption.

full rationale

The derivation chain is self-contained. Assumption (A3) supplies the conditioned Gaussian innovation model as an explicit input; it is not derived from Theorem 5. Achievability (Section V) uses Gaussian signaling with P_n = 2σ₀²√(δ/n), computes D(N(0,σ₀²+P_n)∥N(0,σ₀²)) = (n/2)(ρ_n − ln(1+ρ_n)) ≤ δ via (17)–(19), and then applies Bob's AWGN information bound to obtain B_n = R꜀*√n(1+o(1)). The converse (Section VI) proves Lemma 4 from differential entropy and Hadamard's inequality, converts the relative-entropy budget into ΣP_t² ≤ 4σ₀⁴δ, and combines Cauchy–Schwarz with the AWGN bound. The constant σ₀² is a stated model parameter, not a fitted quantity relabeled as a prediction. The paper's citations of prior SRL work are used only for the standard relative-entropy/detectability connection; the Gaussian-innovation premise is supported by standard filtering and cancellation references [33]–[36], not by a self-citation chain or an imported uniqueness theorem. The paper repeatedly and honestly flags that non-Gaussian residuals and non-vanishing variance uncertainty are outside scope (Sections II-A, IV-C, and condition T2), making Theorem 5 a conditional characterization rather than a hidden restatement of its inputs. No load-bearing step reduces to its own definition or to the authors' earlier work.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the posited Gaussian residual model (A3) and the unmeasured residual-floor parameter σ_e². These are not derived from a physical model but are stated as the regime in which the theorem applies; no new physical entities are invented.

free parameters (2)
  • σ_e² (irreducible cancellation-error variance)
    Introduced in Section II-A (A3) as the residual noise after Willie's best block-level cancellation. The main constant R_C* is proportional to σ_0² = σ_W² + σ_e². It is not derived from a physical model; it is a posited parameter that the analysis depends on.
  • σ_{0,t}² (heterogeneous innovation variance profile)
    In Section III-F, the scheduler assumes a known time-varying profile σ_{0,t}², with the optimal power P_t* ∝ σ_{0,t}⁴. The profile is an input, not measured or derived.
assumptions (5)
  • domain assumption (A1) Trackable public component
    Section II-A: the public component must be publicly specified or trackable by Willie from S. Without this, the cancellation residual model is undefined.
  • domain assumption (A2) Warden-favorable cancellation
    Section II-A: Willie can choose any physically admissible canceller g; the analysis is conditional on the induced residual.
  • domain assumption (A3) Known Gaussian innovation residual
    Section II-A: the post-cancellation null residual is zero-mean Gaussian with known variance; R_0 = N(s), R_1 = X_c + N(s) with N(s) independent of X_c. This is the core assumption; the theorem fails without it.
  • domain assumption (A4) No hidden-resource public-component design
    Section II-A: Alice cannot hide information in untrackable features of the public component.
  • domain assumption Bob-side overt removability
    Section II-A: Bob can estimate/remove U_B^n with residual negligible on the √n scale.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Square-Root Law for Covert Communication with Warden-Favorable Side Information." pith.science (2026). https://pith.science/paper/GFFX4AQA

@misc{pith2026260714013,
  author       = {Pith},
  title        = {Pith review of: Square-Root Law for Covert Communication with Warden-Favorable Side Information},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GFFX4AQA}},
  note         = {Machine review of arXiv:2607.14013}
}
abstract

Covert communication enables Alice to transmit to Bob while making the transmission difficult for Willie to detect. We study a scalar Gaussian covert-overlay model in which Alice's low-power covert signal is superimposed on an aggregate public component generated by Alice or other trackable sources. Willie is given all physically obtainable side information, including protocol details, timing, pilots, channel estimates, and calibration information, and subtracts his best estimate of the public component before testing. Covertness is imposed on the resulting residual through a relative-entropy constraint with budget $\delta$ conditioned on Willie's side information. In the stationary case, the residual under no covert transmission has variance $\sigma_0^2=\sigma_W^2+\sigma_e^2$, where $\sigma_W^2$ is Willie's receiver-noise variance and $\sigma_e^2$ is the irreducible cancellation error. Over $n$ channel uses, the maximal reliably transmissible covert payload is $R_C^\star\sqrt{n}(1+o(1))$ bits, where $R_C^\star=\frac{\sigma_0^2}{\sigma_B^2\ln 2}\sqrt{\delta}$, and $\sigma_B^2$ is Bob's receiver-noise variance. Thus, the square-root-law (SRL) constant is governed by the variance at Willie's actual detector input, not by receiver noise alone. Low-power Gaussian signaling achieves this constant, and a matching converse establishes first-order optimality within the conditioned additive Gaussian innovation model. For known time-varying conditioned residual variances, we also derive the first-order allocation, which assigns more covert power to larger residual variances. The results require a Gaussian post-cancellation null residual with known conditioned variance; non-Gaussian residuals and fixed non-vanishing variance uncertainty are outside the scope of this paper.

Figures

Figures reproduced from arXiv: 2607.14013 by the authors.

Figure 1
Figure 1. Warden-favorable AWGN model under the conditioned post-subtraction relative-entropy constraint. Alice transmits the [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 4
Figure 4. shows that increasing the residual floor raises the SRL constant while preserving the same asymptotic conver￾gence behavior. D. Heterogeneous Innovation Variances: The σ 4 Scheduler For time-varying conditioned innovation variances, the small-signal relative-entropy budget is Pn t=1 P 2 t 4σ 4 0,t ≲ δ. Max￾imizing the first-order total covert power under this constraint gives Pt ∝ σ 4 0,t. For a known admissible var… view at source ↗
Figure 5
Figure 5. σ 4 -based versus uniform allocation for log-normal {σ 2 0,t}, with n = 3 × 104 , σ 2 W = 2, σ 2 e = σ 2 W , log-spread 0.6, and δ = 1. E. Post- vs. Pre-Residualization Relative-Entropy Constraints [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figures from the paper (1 more)
Figure 6
Figure 6. Figure 6: Post- versus pre-residualization design for [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 1 canonical work pages

  1. [1]

    Square root law for com- munication with low probability of detection on AWGN channels,

    B. A. Bash, D. Goeckel, and D. Towsley, “Square root law for com- munication with low probability of detection on AWGN channels,” in Proc. IEEE Int. Symp. Inf. Theory (ISIT), Cambridge, MA, USA, 2012, pp. 448–452, doi: 10.1109/ISIT.2012.6284228

    arXiv 2012

  2. [2]

    Limits of reliable commu- nication with low probability of detection on AWGN channels,

    B. A. Bash, D. Goeckel, and D. Towsley, “Limits of reliable commu- nication with low probability of detection on AWGN channels,”IEEE J. Sel. Areas Commun., vol. 31, no. 9, pp. 1921–1930, Sept. 2013, doi: 10.1109/JSAC.2013.130923

    arXiv 1921

  3. [3]

    Fundamental limits of commu- nication with low probability of detection,

    L. Wang, G. W. Wornell, and L. Zheng, “Fundamental limits of commu- nication with low probability of detection,”IEEE Trans. Inf. Theory, vol. 62, no. 6, pp. 3493–3503, Jun. 2016, doi: 10.1109/TIT.2016.2548471

    arXiv 2016

  4. [4]

    Covert communication over noisy channels: A resolvabil- ity perspective,

    M. R. Bloch, “Covert communication over noisy channels: A resolvabil- ity perspective,”IEEE Trans. Inf. Theory, vol. 62, no. 5, pp. 2334–2354, May 2016, doi: 10.1109/TIT.2016.2530089

    arXiv 2016

  5. [5]

    First- and second-order asymptotics in covert communication,

    M. Tahmasbi and M. R. Bloch, “First- and second-order asymptotics in covert communication,”IEEE Trans. Inf. Theory, vol. 65, no. 4, pp. 2190–2212, Apr. 2019, doi: 10.1109/TIT.2018.2878526

    arXiv 2019

  6. [6]

    Time-division is optimal for covert communication over some broadcast channels,

    V . Y . F. Tan and S.-H. Lee, “Time-division is optimal for covert communication over some broadcast channels,” inProc. IEEE Inf. Theory Workshop (ITW), Guangzhou, China, 2018, pp. 1–5, doi: 10.1109/ITW.2018.8613413

    arXiv 2018

  7. [7]

    Covert communication over a K-user multiple-access channel,

    K. S. K. Arumugam and M. R. Bloch, “Covert communication over a K-user multiple-access channel,”IEEE Trans. Inf. Theory, vol. 65, no. 11, pp. 7020–7044, Nov. 2019, doi: 10.1109/TIT.2019.2930484

    arXiv 2019

  8. [8]

    Fundamental limits of covert commu- nication over MIMO AWGN channel,

    A. Abdelaziz and C. E. Koksal, “Fundamental limits of covert commu- nication over MIMO AWGN channel,” inProc. IEEE Conf. Commun. Netw. Security (CNS), Las Vegas, NV , USA, 2017, pp. 1–9, doi: 10.1109/CNS.2017.8228657

    arXiv 2017

Show all 36 references
  1. [9]

    Covert communication gains from adversary’s ignorance of transmission time,

    B. A. Bash, D. Goeckel, and D. Towsley, “Covert communication gains from adversary’s ignorance of transmission time,”IEEE Trans. Wireless Commun., vol. 15, no. 12, pp. 8394–8405, Dec. 2016, doi: 10.1109/TWC.2016.2614502

    2016

  2. [10]

    Covert communication in the presence of an uninformed jammer,

    T. V . Sobers, B. A. Bash, S. Guha, D. Towsley, and D. Goeckel, “Covert communication in the presence of an uninformed jammer,”IEEE Trans. Wireless Commun., vol. 16, no. 9, pp. 6193–6206, Sept. 2017, doi: 10.1109/TWC.2017.2720736

    2017

  3. [11]

    Covert wireless communication with artificial noise generation,

    R. Soltani, D. Goeckel, D. Towsley, B. A. Bash, and S. Guha, “Covert wireless communication with artificial noise generation,”IEEE Trans. Wireless Commun., vol. 17, no. 11, pp. 7252–7267, Nov. 2018, doi: 10.1109/TWC.2018.2865946

    2018

  4. [12]

    Covert communication with the help of relay and channel uncertainty,

    J. Wang, W. Tang, Q. Zhu, X. Li, H. Rao, and S. Li, “Covert communication with the help of relay and channel uncertainty,”IEEE Wireless Commun. Lett., vol. 8, no. 1, pp. 317–320, Feb. 2019, doi: 10.1109/LWC.2018.2872058

    2019

  5. [13]

    On covert communication with noise uncertainty,

    B. He, S. Yan, X. Zhou, and V . K. N. Lau, “On covert communication with noise uncertainty,”IEEE Commun. Lett., vol. 21, no. 4, pp. 941– 944, Apr. 2017, doi: 10.1109/LCOMM.2016.2647716

    2017

  6. [14]

    Covert communication in fading channels under channel uncertainty,

    K. Shahzad, X. Zhou, and S. Yan, “Covert communication in fading channels under channel uncertainty,” inProc. IEEE Veh. Technol. Conf. (VTC Spring), Sydney, NSW, Australia, 2017, pp. 1–5, doi: 10.1109/VTCSpring.2017.8108525

    2017

  7. [15]

    Achieving positive rate with undetectable communication over AWGN and Rayleigh channels,

    S. Lee and R. J. Baxley, “Achieving positive rate with undetectable communication over AWGN and Rayleigh channels,” inProc. IEEE Int. Conf. Commun. (ICC), Sydney, NSW, Australia, 2014, pp. 780–785, doi: 10.1109/ICC.2014.6883414

    2014

  8. [16]

    Achieving positive rate with undetectable communication over MIMO Rayleigh channels,

    S. Lee, R. J. Baxley, J. B. McMahon, and R. S. Frazier, “Achieving positive rate with undetectable communication over MIMO Rayleigh channels,” inProc. IEEE Sensor Array Multichannel Signal Pro- cess. Workshop (SAM), A Coruna, Spain, 2014, pp. 257–260, doi: 10.1109/SAM.2014.6882389

    2014

  9. [17]

    Covert communication with the help of an uninformed jammer achieves positive rate,

    T. V . Sobers, B. A. Bash, D. Goeckel, S. Guha, and D. Towsley, “Covert communication with the help of an uninformed jammer achieves positive rate,” inProc. Asilomar Conf. Signals, Syst., Comput., Pacific Grove, CA, USA, 2015, pp. 625–629, doi: 10.1109/ACSSC.2015.7421206

    2015

  10. [18]

    Effective secrecy: Reliability, confusion and stealth,

    J. Hou and G. Kramer, “Effective secrecy: Reliability, confusion and stealth,” inProc. IEEE Int. Symp. Inf. Theory (ISIT), Honolulu, HI, USA, 2014, pp. 601–605, doi: 10.1109/ISIT.2014.6874903

    2014

  11. [19]

    Effective secrecy: Reliability, confusion, and stealth,

    J. Hou, G. Kramer, and M. Bloch, “Effective secrecy: Reliability, confusion, and stealth,” inInformation Theoretic Security and Privacy of Information Systems, R. F. Schaefer, H. Boche, A. Khisti, and H. V . Poor, Eds. Cambridge, U.K.: Cambridge Univ. Press, 2017, pp. 3–20

    2017

  12. [20]

    Covert communications over multiple overt channels,

    S. W. Kim and H. Q. Ta, “Covert communications over multiple overt channels,”IEEE Trans. Commun., vol. 70, no. 2, pp. 1112–1124, Feb. 2022, doi: 10.1109/TCOMM.2021.3127532

    2022

  13. [21]

    Covert communications over non-orthogonal multiple overt channels,

    H. Q. Ta, K. Ho-Van, D. B. da Costa, S. W. Kim, and H. Oh, “Covert communications over non-orthogonal multiple overt channels,” IEEE Access, vol. 10, pp. 122361–122375, 2022, doi: 10.1109/AC- CESS.2022.3223357

    2022

  14. [22]

    Rate-splitting multiple access for covert communica- tions,

    Y . Zhanget al., “Rate-splitting multiple access for covert communica- tions,”IEEE Wireless Commun. Lett., vol. 13, no. 6, pp. 1685–1689, Jun. 2024, doi: 10.1109/LWC.2024.3386548

    2024

  15. [23]

    Covert multi-access communication with a non-covert user,

    A. Bounhar, M. Sarkiss, and M. Wigger, “Covert multi-access communication with a non-covert user,” inProc. IEEE Int. Conf. Commun. (ICC), Denver, CO, USA, 2024, pp. 1473–1478, doi: 10.1109/ICC51166.2024.10622567

    2024

  16. [24]

    Whispering secrets in a crowd: Leveraging non-covert users for covert communications,

    A. Bounhar, M. Sarkiss, and M. Wigger, “Whispering secrets in a crowd: Leveraging non-covert users for covert communications,”arXiv preprint arXiv:2408.12962, 2024

    2024 arXiv

  17. [25]

    Covert com- munication and secure transmission over untrusted relaying networks in the presence of multiple wardens,

    M. Forouzesh, P. Azmi, A. Kuhestani, and P. L. Yeoh, “Covert com- munication and secure transmission over untrusted relaying networks in the presence of multiple wardens,”IEEE Trans. Commun., vol. 68, no. 6, pp. 3737–3749, Jun. 2020, doi: 10.1109/TCOMM.2020.2978206

    2020

  18. [26]

    Covert com- munication achieved by a greedy relay in wireless networks,

    J. Hu, S. Yan, X. Zhou, F. Shu, J. Li, and J. Wang, “Covert com- munication achieved by a greedy relay in wireless networks,”IEEE Trans. Wireless Commun., vol. 17, no. 7, pp. 4766–4779, Jul. 2018, doi: 10.1109/TWC.2018.2831217

    2018

  19. [27]

    Wireless covert communications aided by distributed cooperative jamming over slow fading channels,

    T.-X. Zheng, Z. Yang, C. Wang, Z. Li, J. Yuan, and X. Guan, “Wireless covert communications aided by distributed cooperative jamming over slow fading channels,”IEEE Trans. Wireless Commun., vol. 20, no. 11, pp. 7026–7039, Nov. 2021, doi: 10.1109/TWC.2021.3080382

    2021

  20. [28]

    Collaborative secret and covert communications for multi-user multi-antenna uplink UA V systems: Design and optimization,

    J. Xu, L. Bai, X. Xie, and L. Zhou, “Collaborative secret and covert communications for multi-user multi-antenna uplink UA V systems: Design and optimization,”IEEE Trans. Wireless Commun., vol. 24, no. 7, pp. 6020–6035, Jul. 2025, doi: 10.1109/TWC.2025.3550911

    2025

  21. [29]

    Covert communications: A comprehensive survey,

    X. Chenet al., “Covert communications: A comprehensive survey,” IEEE Commun. Surveys Tuts., vol. 25, no. 2, pp. 1173–1198, 2nd Quart., 2023, doi: 10.1109/COMST.2023.3263921

    2023

  22. [30]

    Hiding informa- tion in noise: Fundamental limits of covert wireless communication,

    B. A. Bash, D. Goeckel, D. Towsley, and S. Guha, “Hiding informa- tion in noise: Fundamental limits of covert wireless communication,” IEEE Commun. Mag., vol. 53, no. 12, pp. 26–31, Dec. 2015, doi: 10.1109/MCOM.2015.7355562

    2015

  23. [31]

    Simultaneous secure and covert transmissions against two attacks under practical assumptions,

    M. Forouzesh, F. S. Khodadad, P. Azmi, A. Kuhestani, and H. Ahmadi, “Simultaneous secure and covert transmissions against two attacks under practical assumptions,”IEEE Internet Things J., vol. 10, no. 12, pp. 10160–10171, Jun. 15, 2023, doi: 10.1109/JIOT.2023.3237640

    2023

  24. [32]

    Covert communi- cation with channel-state information at the transmitter,

    S.-H. Lee, L. Wang, A. Khisti, and G. W. Wornell, “Covert communi- cation with channel-state information at the transmitter,”IEEE Trans. Inf. Forensics Security, vol. 13, no. 9, pp. 2310–2319, Sept. 2018, doi: 10.1109/TIFS.2018.2818650

    2018

  25. [33]

    B. D. O. Anderson and J. B. Moore,Optimal Filtering. Englewood Cliffs, NJ, USA: Prentice-Hall, 1979

    1979

  26. [34]

    Analysis of a simple successive interference cancellation scheme in a DS/CDMA system,

    P. Patel and J. Holtzman, “Analysis of a simple successive interference cancellation scheme in a DS/CDMA system,”IEEE J. Sel. Areas Com- mun., vol. 12, no. 5, pp. 796–807, Jun. 1994, doi: 10.1109/49.298053

    1994 doi

  27. [35]

    Asymptotically optimal nonlinear MMSE multiuser detection based on multivariate Gaussian approxima- tion,

    P. H. Tan and L. K. Rasmussen, “Asymptotically optimal nonlinear MMSE multiuser detection based on multivariate Gaussian approxima- tion,”IEEE Trans. Commun., vol. 54, no. 8, pp. 1427–1438, Aug. 2006, doi: 10.1109/TCOMM.2006.878829

    2006

  28. [36]

    Capacity of the Gaus- sian two-hop full-duplex relay channel with residual self-interference,

    N. Zlatanov, E. Sippel, V . Jamali, and R. Schober, “Capacity of the Gaus- sian two-hop full-duplex relay channel with residual self-interference,” IEEE Trans. Commun., vol. 65, no. 3, pp. 1005–1021, Mar. 2017, doi: 10.1109/TCOMM.2017.2653112

    2017

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.