{"id":"68da4799-8104-4fbf-8ebb-74b108d9b10f","arxiv_id":"2507.08293","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The 2D ambiguity function of continuous-time AFDM chirp subcarriers is derived, showing a periodic comb with parallelogram unambiguity regions and a Doppler shift equal to the subcarrier frequency difference for cross-ambiguity.","lead":"The paper derives the range and velocity sensing properties of AFDM, a chirp-based wireless waveform proposed for 6G. It shows the waveform's ambiguity function forms a repeating pattern with a parallelogram-shaped unambiguous detection region, and that guard symbols enable interference-free sensing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The global pulse-lattice and guard-band conclusions rest on a closed-form derivation that covers only a representative delay interval (eq. 18), which for m=0 has zero length; the promised 'minor modifications' for all remaining cases are never shown.","rationale":"I agree with the reader's assessment: the weakest point is the unproved extrapolation from a representative delay interval to the full global structure. This is not a mere cosmetic omission because the global pulse lattice, the parallelogram, and the guard-band region are the basis for the paper's practical conclusions. The concern is about completeness of proof, not a demonstrated falsehood; simulations support the shown regimes and there is no evidence of circularity or data fitting. I considered the pilot-scaling issue in Eq. (52) (the pilot component should scale as xp, not |xp|^2, given the definition of APD), but that is a localized inconsistency that affects the PDR discussion rather than the core lattice structure. The most load-bearing point remains the missing derivation for omitted delay intervals. The recommended verdict is unchanged: CONDITIONAL, pending the exhaustive check.","tokens_in":22598,"tokens_out":9867,"duration_ms":118163,"concrete_test":"Directly evaluate (13) by numerical quadrature on a fine grid (delay step Δt/4, Doppler step Δf/4) over τ∈[0,T] for N=512, C=13, m=0 and m=47 (and an m near N/2, where the covered interval is largest), using the exact piecewise φm(t) from (4)–(5). Extract all local maxima of |A|^2 above 0.5 of the global maximum and compare their (τ,ν) coordinates with the predicted lattice generated by (Δt,CΔf) and the TSC-spaced parallel lines of Remark 2. If every predicted pulse is present and no extras appear, especially for m=0 and for τ outside (18), the extrapolation is sound; otherwise Proposition 1 and the guard-band conclusions need re-derivation before acceptance.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central structural claims—Property 2's periodic-like grid, Remark 1's Δt/CΔf pulse spacing, Corollary 1's unit-area parallelogram, Proposition 1's unambiguous sensing region, and the Sec. V-B guard-band interference-free region—are all justified from the closed-form AAF/CAF expressions (21)/(44). But those expressions are derived only under τ ∈ [kTSC, min{(TSC−tm,1), tm,1}+kTSC] for the AAF and the analogous interval (40) plus m>m1 for the CAF. This interval has length min{mΔt/C, TSC−mΔt/C} ≤ TSC/2, and for m=0 it degenerates to the single point τ=kTSC. The paper states (Sec. III-A) that 'due to space limitations, we only demonstrate one representative case, from which the remaining cases can be obtained with minor modifications,' but none of those cases, and no symmetry argument, is provided. If the omitted intervals yield different summation limits that change the constructive-interference condition (e.g., different effective q shifts or a different ddelay), the pulse grid and the parallelogram could shift or acquire extra pulses, which would invalidate Proposition 1 and the interference-free parallelogram. Numerical simulations in Fig. 10 cover only selected (m,C) and do not exhaust the delay interval; they provide support but not proof. Eq. (52)'s pilot scaling is a separate typo-level inconsistency (the pilot term should scale as xp, not |xp|^2), but it is secondary to the unproved extrapolation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes the aperiodic ambiguity functions of continuous-time AFDM signals for integrated sensing and communications. The authors derive closed-form expressions for the auto-ambiguity function (AAF) of a single AFDM chirp subcarrier and for the cross-ambiguity function (CAF) between two different chirp subcarriers. From these expressions they identify a 'spike-like' local pulse with delay and Doppler resolutions approximately Δt and Δf, and a 'periodic-like' global grid of pulses with adjacent spacings Δt and CΔf, each parallelogram having unit area. They then define an unambiguity parallelogram and argue that choosing the chirp parameter c1 (equivalently C) so that all target delay-Doppler shifts lie inside one parallelogram enables unambiguous monostatic sensing. For bistatic sensing they analyze the CAF between a pilot subcarrier and an AFDM frame with pilot, guard, and data symbols, and show that inserting guard symbols creates an interference-free parallelogram. The theoretical results are supported by numerical simulations for selected parameter values.","tokens_in":22895,"tokens_out":2472,"duration_ms":28225,"significance":"If the central structural claims are fully established, the paper provides a useful characterization of AFDM's sensing capabilities and a concrete design guideline for choosing c1 to avoid ambiguity in AFDM-ISAC systems. The derivations start directly from the signal definition and the ambiguity-function integral (12), with no fitted parameters, and the numerical simulations in Sec. VI are consistent with the claimed local and global pulse structure for the cases shown. The comparison with OFDM, SCM, and OTFS subcarriers also gives useful insight into why AFDM chirp subcarriers offer simultaneous delay and Doppler perceptibility. However, the central claims are currently justified only for a restricted delay interval, and the promised extension to all other delay cases is not provided; this gap affects the global pulse-lattice property, the unambiguity parallelogram, and the guard-band interference-free region. With the missing cases supplied, the results would be a solid contribution to waveform design for ISAC.","major_comments":[{"comment":"The closed-form AAF in Eq. (21) is derived only for the delay interval τ ∈ [kTSC, min{(TSC − tm,1), tm,1} + kTSC], which has length at most TSC/2 and degenerates to a single point for m = 0. The text states 'due to space limitations, we only demonstrate one representative case, from which the remaining cases can be obtained with minor modifications,' but no such cases or modifications are shown, and no symmetry argument is given. The pulse-lattice structure of Property 2, the pulse spacings in Remark 1, and Corollary 1 all rely on the summation limits and the constructive-interference conditions derived from this one interval. If a different delay interval yields different effective values of qm(t) − qm(t − τ) or different residue phase terms, the pulse grid could shift, acquire extra pulses, or change spacing, which would invalidate the unambiguity parallelogram and Proposition 1. The authors should either derive the closed-form expressions for the remaining delay intervals or provide a rigorous argument that the representative case is sufficient.","section":"Sec. III-A, Eq. (18)-(21)"},{"comment":"The CAF derivation suffers from the same restricted-interval problem as the AAF: Eq. (44) is derived only under τ ∈ [kTSC, min{(TSC − tm1,1), tm,1} + kTSC], along with the condition m > m1. For many subcarrier pairs the interval defined by min{(TSC − tm1,1), tm,1} may be empty, and the paper does not explain how the decomposition in Eq. (43) behaves in those cases. Proposition 2's claim of an additional Doppler shift δf^{m,m1} is obtained from the FA condition in the representative case, and it is not clear that the same shift applies in all other delay intervals. Since the bistatic sensing results in Sec. V-B and the guard-band conclusions depend on the CAF pulse locations, the omission affects load-bearing conclusions beyond Proposition 2 itself.","section":"Sec. IV, Eq. (40)-(44)"},{"comment":"In Eq. (52), the pilot component of A_{s,φ_{mp}}(τ,ν) is written as |x_p|^2 A_{φ_{mp},φ_{mp}}(τ,ν). According to the definition A_{a,b}(τ,ν) = ∫ a(t) b*(t−τ) e^{-j2πνt} dt, the pilot term should be x_p A_{φ_{mp},φ_{mp}}(τ,ν), not |x_p|^2 times the AAF, because the AAF of the pilot subcarrier already contains the full pilot amplitude and the cross-term between the pilot and data symbols is x_p^* Σ x[m] A_{φ_m,φ_{mp}}(τ,ν). The squared amplitude appears to be an error that changes the relative levels of the pilot and data components and the interpretation of the PDR. This should be corrected and the subsequent remarks checked for consistency.","section":"Sec. V-B, Eq. (52)"}],"minor_comments":[{"comment":"The word 'Subsitituting' in the sentence before Eq. (10) is a typo and should read 'Substituting'.","section":"Sec. II-C"},{"comment":"In the discussion following Eq. (30), the expression 'PC−˜qm(τ ) i=0 S i b' appears with a formatting error; it should read 'Σ_{i=0}^{C−q̃_m(τ)} S_i^b'.","section":"Sec. III-B"},{"comment":"The word 'visuallized' in the proof of Proposition 2 should be 'visualized'.","section":"Sec. IV"},{"comment":"The ratio ρ for the EP structure is reported as 45.5%; the manuscript should clarify whether this accounts for both the pilot and guard symbols and should state the number of guard symbols Q used in the figure, since Q is introduced formally just before.","section":"Sec. V-B, Fig. 8"},{"comment":"The text mentions 'NOTFS and NOTFS denote the numbers of Doppler and delay bins'; one of these should be MOTFS, and the sentence should be corrected.","section":"Sec. VI"},{"comment":"The caption says '265-QAM' but the text likely intends '256-QAM'; please verify and correct.","section":"Sec. VI, Fig. 14"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely publishable after the missing delay-interval cases are supplied or a rigorous symmetry argument is provided. The representative-case derivation is a conscious abbreviation, but it currently covers less than half of each subchirp delay range and, for m=0, a zero-length interval; this is not merely a presentation issue because the global pulse-lattice and guard-band conclusions are load-bearing. The Eq. (52) pilot scaling error is also straightforward to fix. I would not reject the manuscript on these grounds, but the revision must address them explicitly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is a genuine result, not a repackaging. The full 2D continuous-time AAF of AFDM chirp subcarriers—rotated periodic pulse grid, adjacent pulse spacings Delta_t and C*Delta_f, unit-area unambiguity parallelogram—and the CAF Doppler shift equal to the subcarrier frequency difference are new relative to the zero-Doppler discrete analysis in [45]. The derivation starts from the signal definition and the AF integral, has no fitted parameters, and the simulations in Figs. 10–14 match the shown regimes, including the random-data averaging. That is solid, reproducible work, and I would not be surprised if it becomes a standard reference for AFDM AF properties.\n\nThe soft spot is exactly where the reader and stress-test point. The closed forms (21) and (44) are derived only on the delay intervals (18) and (40). Those intervals are at most half a subchirp, and for m = 0 the interval degenerates to a single point. The paper says, in Sec. III-A, that due to space limitations only one representative case is demonstrated and the rest follow by minor modifications—but those modifications are never shown. Property 2, Remark 1, Corollary 1, Proposition 1, and the guard-band interference-free region all inherit this unproved extrapolation. The figures suggest the pulse grid is stable across different m (Fig. 10a,b), but pictures are not proof, and if the omitted delay intervals change the summation limits, the grid or parallelogram could shift. This is a genuine gap, though I think the conclusion is probably correct. The fix is straightforward: fill in the remaining intervals, or give a symmetry/periodicity argument reducing them to the representative case.\n\nThe Eq. (52) scaling issue is minor and typo-level: the pilot component should scale as x_p, not |x_p|^2, unless the pilot amplitude is silently folded into the matched-filter definition. Worth cleaning up, but not a threat to the thesis.\n\nWho this is for: anyone in ISAC waveform design, especially 6G AFDM work. The unambiguous-sensing and guard-symbol guidance is actionable, and the analysis directly informs parameter selection. I would send this to review; it deserves referee time, and the missing derivation is fixable in revision. I would also want a referee to check whether the periodic-like grid really repeats across all delay intervals, because that is the hinge.","headline":"Real new AFDM ambiguity-function analysis with a load-bearing derivation gap: full 2D results shown only for a restricted delay interval, and the global pulse-grid and guard-band conclusions rest on an unshown extrapolation.","tokens_in":23457,"tokens_out":2642,"would_cite":true,"duration_ms":31086,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper derives closed-form ambiguity functions for AFDM chirp subcarriers, showing a spike-like local pulse and a periodic grid of pulses that make delay-Doppler sensing unambiguous when the chirp slope is tuned to cover the targets.","keywords":["AFDM","ambiguity function","integrated sensing and communications","chirp subcarrier","delay-Doppler sensing","guard symbols","bistatic sensing","unambiguous sensing"],"falsifier":"Evaluate $|A_{\\phi_m,\\phi_m}(\\tau,\\nu)|^2$ numerically over the full delay range $\\tau\\in[0,T]$, not only the interval in (18), for a fixed $m$ and $C$, and check whether adjacent pulse spacings remain $\\Delta t$ and $C\\Delta f$ and every parallelogram still has area one; a deviation on any omitted subchirp interval would overturn Proposition 1 and the guard-symbol conclusions.","tokens_in":22378,"feed_emoji":"📡","tokens_out":8404,"duration_ms":84557,"temperature":0.7,"pith_summary":"This paper works out the two-dimensional ambiguity function of AFDM, a chirp-based multicarrier waveform, in continuous time. Its central result is that a single AFDM chirp subcarrier has a spike-like peak with delay resolution about $1/B$ and Doppler resolution about $1/T$, plus a periodic-like lattice of weaker pulses along a rotated delay-Doppler axis. The lattice spacing is $\\Delta t$ in delay and $C\\Delta f$ in Doppler, and each lattice cell is a parallelogram of area one. Because the cell is shaped by the chirp parameter $c_1$, the transmitter can stretch the unambiguity region to cover the expected target delays and Dopplers. The paper also derives the cross-ambiguity function between different subcarriers and shows that guard symbols around a pilot create an interference-free parallelogram for bistatic sensing.","feed_headline":"AFDM chirp grid makes sensing unambiguous","feed_subtitle":"One chirp parameter packs target delays and Dopplers into a unit-area grid, sharpening range and velocity sensing.","key_machinery":"The carrying object is the aperiodic ambiguity function $A_{a,b}(\\tau,\\nu)=\\int_{-\\infty}^{\\infty}a(t)b^*(t-\\tau)e^{-j2\\pi\\nu t}dt$, computed for a single AFDM chirp subcarrier, which the paper models as a piecewise phase-modulated signal made of $C$ wrapped subchirps. Two alignment conditions select the pulse locations: frequency alignment, where the instantaneous frequency difference between $\\phi_m(t)$ and $\\phi_m(t-\\tau)e^{j2\\pi\\nu t}$ vanishes, and residue phase alignment, which requires $\\tau=k\\Delta t$. When both hold, the aligned subchirp pieces accumulate energy and produce a pulse; scanning $(\\tau,\\nu)$ therefore traces lines of slope $2\\tilde{c}_1$, with adjacent pulses separated by $\\Delta t$ and $C\\Delta f$. The parallelogram outlined by four neighboring pulses has area one, and the paper calls this cell the unambiguity parallelogram.","core_discovery":"The paper's central claim is that the aperiodic auto-ambiguity function $A_{\\phi_m,\\phi_m}(\\tau,\\nu)$ of an AFDM chirp subcarrier is not a single ridge but a grid: locally it behaves like a spike with mainlobe widths $\\Delta t=1/B$ in delay and $\\Delta f=1/T$ in Doppler, and globally it repeats pulses along lines of slope $2\\tilde{c}_1$, whose adjacent spacings are $\\Delta t$ and $C\\Delta f$. Frequency alignment and residue phase alignment conditions select the pulse locations, and the parallelogram outlined by four neighboring pulses always has area one. For two different subcarriers, the cross-ambiguity function has exactly the same grid shifted by $\\delta_f=(m-m_1)\\Delta f$ along Doppler, so the Doppler shift equals the subcarrier frequency difference. For full frames, random data average into a single thumbtack peak, while inserting guard symbols around a pilot produces a smaller interference-free parallelogram inside the unambiguity region. The practical payoff is that choosing $C$ via $c_1$ to cover the channel's delay-Doppler spread gives unambiguous monostatic sensing, and guard symbols give interference-free bistatic sensing when the pilot is the only known part of the signal.","pith_inferences":["Because every unambiguity parallelogram has area one in delay-Doppler units, tuning $c_1$ only trades unambiguous delay range against Doppler range; the product of the two ranges is fixed by the AFDM time-bandwidth product.","The AFDM subcarrier grid closely resembles the OTFS subcarrier pulse train, so the same unit-cell reasoning may transfer to other full-resource waveforms; the paper notes the resemblance but does not claim transferability.","Replacing rectangular shaping with a root-raised-cosine filter (the paper's stated future work) could preserve or distort the unit-area parallelogram; computing the AAF under that shaping would show which properties are intrinsic to AFDM and which depend on rectangular pulses."],"forward_implications":["For monostatic sensing, choosing $C$ large enough that all target delay and Doppler shifts lie inside one unambiguity parallelogram prevents echoes from different targets being confused, at the cost of higher $c_1$-related overhead.","A single AFDM chirp subcarrier simultaneously resolves delay at $1/B$ and Doppler at $1/T$, unlike an SCM symbol (delay-only) or an OFDM subcarrier (Doppler-only).","In bistatic sensing with only the pilot known, the CAF's Doppler shift $\\delta_f$ between subcarriers means data subcarriers contribute interference offset by their frequency difference; guard symbols placed around the pilot create an interference-free parallelogram.","The interference-free parallelogram is strictly smaller than the unambiguity parallelogram, so bistatic guard-symbol sensing demands a narrower delay-Doppler spread or more guard overhead.","Random-data AFDM frames without pilots retain a thumbtack-like average AAF, so data-only sensing is feasible at the monostatic receiver."],"supporting_citations":[{"why":"Defines AFDM and its DAFT chirp-subcarrier model, the object whose ambiguity functions are analyzed.","marker":"[15]"},{"why":"Supplies the continuous-time piecewise spectrum-wrapping representation of AFDM subcarriers used in the closed-form derivation.","marker":"[39]"},{"why":"Provides the rectangular pulse-shaping model and design criterion adopted for all subcarriers in the analysis.","marker":"[22]"},{"why":"Gives the OTFS subcarrier structure and AF behavior that the paper compares against and extends to AFDM.","marker":"[12]"},{"why":"Establishes the prior zero-Doppler AFDM AF analysis that this paper generalizes to full two-dimensional continuous-time AFs.","marker":"[45]"}],"fun_headline_variants":["AFDM chirp grid packs targets into unit-area cells","Chirp ambiguity grid enables unambiguous ISAC sensing","AFDM's aperiodic grid sharpens range and Doppler estimation","Unit-area ambiguity cells make AFDM sensing unambiguous"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The closed form is derived for one representative delay interval, (18), and the paper says the remaining cases are obtained with minor modifications; the global grid, unit-area parallelogram, and guard-band conclusions assume those unshown cases behave identically.","fun_headline_variants_meta":{"raw":{"variants":["AFDM chirp grid packs targets into unit-area cells","Chirp ambiguity grid enables unambiguous ISAC sensing","AFDM's aperiodic grid sharpens range and Doppler estimation","Unit-area ambiguity cells make AFDM sensing unambiguous"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000513,"raw_usage":{"total_tokens":2541,"prompt_tokens":1039,"completion_tokens":1502,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":655,"completion_tokens_details":{"reasoning_tokens":1436}},"tokens_in":655,"tokens_out":1502,"duration_ms":11996,"temperature":1.0,"reasoning_tokens":1436,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:23:36.033810+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate $|A_{\\phi_m,\\phi_m}(\\tau,\\nu)|^2$ numerically over the full delay range $\\tau\\in[0,T]$, not only the interval in (18), for a fixed $m$ and $C$, and check whether adjacent pulse spacings remain $\\Delta t$ and $C\\Delta f$ and every parallelogram still has area one; a deviation on any omitted subchirp interval would overturn Proposition 1 and the guard-symbol conclusions.","supporting_citations":[{"cited_title":"Affine frequency division multiplexing for next-generation wireless communications,","cited_arxiv_id":null,"evidence_quote":"Defines AFDM and its DAFT chirp-subcarrier model, the object whose ambiguity functions are analyzed."},{"cited_title":"Integrated sensing and communications with affine frequency division multiplexing,","cited_arxiv_id":null,"evidence_quote":"Supplies the continuous-time piecewise spectrum-wrapping representation of AFDM subcarriers used in the closed-form derivation."},{"cited_title":"Evaluation and design criterion for pulse- shaped AFDM,","cited_arxiv_id":null,"evidence_quote":"Provides the rectangular pulse-shaping model and design criterion adopted for all subcarriers in the analysis."},{"cited_title":"AFDM-based bistatic integrated sensing and communication in static scatterer environments,","cited_arxiv_id":null,"evidence_quote":"Establishes the prior zero-Doppler AFDM AF analysis that this paper generalizes to full two-dimensional continuous-time AFs."}],"review_version":1}