{"id":"1b6c2199-0d94-4f2d-885f-608e96d2cce6","arxiv_id":"1908.09244","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For phase-aligned secure precise wireless transmission, optimizing the antenna amplitude weights with a leakage criterion or with a maximum-receive-power plus null-steering criterion gives lower bit-error rate and higher secrecy rate than equal amplitude beamforming in simulations.","lead":"The authors propose two antenna-amplitude rules, leakage and maximum-receive-power, for a radio system that focuses a secret message at an intended receiver while focusing artificial noise at an eavesdropper. In simulations, both rules improve bit-error rate and secrecy rate compared with the existing equal-amplitude design, with only a small difference between them.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Magnitude projection invalidates the stated SLNR/null-space solutions; Eqs. (5)-(14) do not describe the simulated P|a|/Q|b| transmit vectors.","rationale":"The paper has a plausible core idea: with phase alignment fixed, amplitude weights can improve secrecy, and generalized eigenvector/SVD computations are standard. The external assumption of known LoS phases is a limitation, not an internal flaw. The serious problem is that the derivations in Section III solve complex-domain problems, while the transmitted vectors are built from the magnitudes of those solutions. This is not a mere typo: the Max-RP null constraints and the leakage objective are not preserved by the magnitude projection, so the algorithms as claimed are not the algorithms as simulated. The reader's conditional verdict is appropriate; my concern reinforces it rather than changing it. The paper could be repaired by optimizing directly over real nonnegative amplitudes or by explicitly presenting the magnitude-projection as a heuristic with direct numerical validation, and by fixing the swapped steering vectors and beta in Eq. (9). Because no code or data are provided, an independent regeneration is necessary before the central performance claim is settled.","tokens_in":6462,"tokens_out":11161,"duration_ms":116431,"concrete_test":"At the paper's simulation parameters (N=32, theta_B=30 deg, R_B=650 m, theta_E=100 deg, R_E=550 m, SNR=14 dB), compute a from Eq. (7) and b from the corrected leakage objective, form v_CM=P|a| and v_AN=Q|b|, and evaluate the normalized residual r=|h_E^H v_CM|/|h_B^H v_CM| for the Max-RP CM design and the analogous Bob residual for the AN design. If r is not below -40 dB, the null constraint is violated and the magnitude projection invalidates the optimization. Then run one Monte Carlo BER/SR point comparing the magnitude-projected vectors with the true complex solutions; a material change in the curves would show the simulations do not correspond to the derived algorithms.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing concern is internal. Section III optimizes complex vectors a and b (Eqs. (5)-(14)), but the final transmit vectors are P|a| and Q|b|, because the AB vector is set to [|a(0)|,...,|a(N-1)|]^T. Magnitude projection is not a feasible point of the stated problems. For Max-RP, Eq. (11) imposes h_E^H P a = 0; after projection the residual h_E^H P|a| = (1/sqrt(N)) sum_n exp(j(Psi_n(theta_B,R_B)-Psi_n(theta_E,R_E))) |a_n| is generally nonzero, so the promised null toward Eve is lost. The same failure nulls the AN toward Bob via Eq. (13). For leakage, the complex generalized eigenvector in Eqs. (7)/(10) does not maximize SLNR after its phases are discarded; the amplitude-only optimum is a different vector. Consequently, the received-signal models in Eqs. (3)-(4), which use a and b, do not describe the simulated vectors P|a| and Q|b|. In addition, Eq. (9) has the Bob/Eve steering vectors swapped and the noise term uses beta instead of (1-beta), making the AN leakage objective as printed inconsistent. With no code or numerical tables, the reported BER/SR advantage in Figs. 3-4 cannot be independently reproduced from the derivations.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a secure precise wireless transmission/precise jamming scheme, RSS-PJ-AN, in which the transmit beamforming vectors for the confidential message (CM) and artificial noise (AN) are designed by separating phase alignment from amplitude beamforming. The phase parts are fixed to align, respectively, with the legitimate receiver Bob and the eavesdropper Eve, and the amplitude parts are optimized by two proposed rules: a leakage-based rule (maximizing SLNR) and a Max-RP rule (maximizing received power at the desired receiver subject to a null constraint at the other receiver). The paper claims, on the basis of Monte-Carlo-style simulations in Section IV, that both proposed amplitude beamforming schemes outperform the equal-amplitude baseline EAB in bit-error-rate and secrecy-rate at medium and high SNR, while performing similarly at low SNR. The central claimed contribution is the ability to reduce the joint phase-and-amplitude design to a much simpler amplitude-only problem under phase alignment.","tokens_in":6715,"tokens_out":5069,"duration_ms":57284,"significance":"If the claim were fully substantiated, the decoupling of phase alignment and amplitude optimization would be a useful and low-complexity design principle for secure precise wireless transmission, extending SPWT ideas to precise jamming. The optimization objectives used in Section III (SLNR and null-space receive power) are not the same as the reported evaluation metrics (BER and secrecy rate), so the paper is not circularly fitting its targets. The proposed solutions are closed-form (generalized eigenvectors and SVD-based null-space projections), which is an attractive feature for implementation. However, in its current form the paper does not deliver a sound derivation of the simulated transmit vectors: the amplitude-only vectors obtained after magnitude projection are not feasible points of the complex optimization problems solved in Section III, and the exact nulls promised by the Max-RP rule are generally destroyed by that projection. The paper also contains an apparent swapping of Bob/Eve steering vectors in Eq. (9).","major_comments":[{"comment":"The optimization is performed over complex vectors a and b, but the final transmit vectors are P|a| and Q|b|. This is stated explicitly in Section II and again after Eqs. (7), (10), (12), and (14): the element magnitudes of vCM and vAN are set to |a(n)| and |b(n)|. The received-signal models in Eqs. (3)–(4) and the objectives in Eqs. (6), (9), (11)–(14), however, all use the full complex vectors Pa and Qb. Magnitude projection is not a feasible point of these optimization problems. In particular, for Max-RP the constraint h_E^H P a = 0 in Eq. (11) does not imply h_E^H P|a| = 0; the residual 1/sqrt(N) sum_n exp(j(Psi_n(theta_B,R_B)-Psi_n(theta_E,R_E))) |a_n| is generally nonzero, so the claimed null toward Eve is lost. The same issue invalidates the AN null toward Bob in Eq. (13). For the leakage-based scheme, the generalized eigenvector in Eqs. (7) and (10) does not maximize the stated SLNR after its phases are discarded; the amplitude-only optimum is a different vector. The derivation in Section III must be redone for the actual simulated vectors, either by optimizing directly over real nonnegative amplitude vectors or by analyzing the projected vectors; as written, the received-signal models used in the simulation are not those derived.","section":"Eq. (9)"},{"comment":"The SLNR objective for the AN vector b is internally inconsistent as printed. Since the AN is intended as the useful signal at Eve, the numerator should be proportional to |h_E^H Q b|^2, and the leakage term in the denominator should be proportional to |h_B^H Q b|^2; instead, Eq. (9) places h(theta_B,R_B) in the numerator and h(theta_E,R_E) in the denominator. The noise term also uses beta P_s although the AN power is (1-beta) P_s, so the generalized eigenproblem in Eq. (10) does not solve the SLNR problem described in the text. This must be corrected before the leakage-based scheme can be evaluated, since the optimal b is currently obtained from a different objective than the stated one.","section":"Section IV, Figs. 2–4"},{"comment":"The central performance claim rests entirely on the simulation figures, but the manuscript gives no code, no data tables, no number of Monte-Carlo trials, and no confidence intervals or error bars. Given the mismatch between the derived vectors and the simulated transmit vectors identified above, it is not possible to verify that the BER and SR curves in Figs. 3 and 4 correspond to the systems analyzed in Section III. The authors should provide either reproducible code, numerical tables, or at least a precise statistical description of the simulations, including the number of independent channel/symbol realizations, so that the claimed one-order-of-magnitude BER improvement can be independently checked.","section":"Section II, after Eq. (1)"},{"comment":"The schemes require exact knowledge of the steering phases to both Bob and Eve, as stated in the assumption of LoP channels and high-resolution DOA estimation, but no sensitivity analysis is provided. Since phase alignment is mandatory for the entire design and the amplitude rules are derived under perfect phase alignment, the practical value of the proposed methods is not established for any phase error. At minimum, the authors should add a robustness study with mismatched angles or ranges, or explicitly state that perfect CSI/DOA is assumed and discuss the expected degradation qualitatively.","section":"Section II, after Eq. (1)"}],"minor_comments":[{"comment":"The notation P = diag[arg(vCM(0)), ..., arg(vCM(N-1))] is formally incorrect: the diagonal entries must be the complex unit-modulus phase factors exp(j arg(vCM(n))), not the real scalar phases. As written, P a would have scalar phase angles as coefficients rather than phase rotations, and the received-signal equations would not follow.","section":"Eq. (2)"},{"comment":"The term 'LoP' appears to be a typo for 'LoS' (line-of-sight); please correct this in the text and in the notation description.","section":"Throughout"},{"comment":"The conclusion and abstract use 'EA' once instead of 'EAB' (equal amplitude beamforming); please make the abbreviation consistent throughout.","section":"Abstract and Section V"},{"comment":"It would help to define explicitly that vCM and vAN are normalized to unit norm, since the optimization constraints a^H a = 1 and b^H b = 1 are used in Section III but the relation between these constraints and the normalization of vCM and vAN is not stated.","section":"Section II, Eq. (2)"},{"comment":"The text says the proposed methods outperform EAB 'particularly in the high SNR region' and describes Fig. 2 as showing performance surfaces, but the figure caption and the amount of detail in the curves are difficult to follow; labeling the curves or providing numerical markers would improve clarity.","section":"Section IV"}],"recommendation":"major_revision","confidential_remarks":"The paper has a potentially interesting idea, but the current manuscript is not publishable in its present form. The load-bearing mismatch between the complex optimization variables and the magnitude-projected transmit vectors, together with the apparent swap in Eq. (9), means that the derived solutions do not describe the simulated systems. This is fixable with a substantial revision that either re-does the optimization over real amplitudes or analyzes the projected vectors, but it is more than a presentation issue. I would ask the authors to address these points before any further consideration."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the move to split phase alignment from amplitude optimization in the SPWT/PJ setup, and then to run simple eigenvector or null-space rules on the amplitude part. That is a sensible thing to try, and the two proposed schemes are concrete and easy to describe. The paper also compares against an equal-amplitude baseline and shows plausible-looking gains in BER and secrecy rate. I believe the authors when they say the simulations showed those gains; I just cannot verify them from the manuscript.\n\nThe soft spot is load-bearing. Section III optimizes complex vectors a and b, but the transmit vectors are P|a| and Q|b| — the phase matrices P and Q times the elementwise magnitudes. A complex eigenvector that maximizes SLNR, or a vector that satisfies the null constraint h_E^H P a = 0, does not keep those properties after you throw away its phases. The magnitude projection is not a feasible point of the stated problems. So the claimed null toward Eve for Max-RP and the claimed SLNR optimality for the leakage scheme do not actually hold for the vectors that are transmitted. The received-signal models in Eqs. (3)–(4) use a and b, not |a| and |b|, so the derivations and the simulations are about different objects.\n\nThere are also smaller internal inconsistencies. Equation (9) has the Bob and Eve steering vectors swapped relative to the stated objective, and the noise term uses beta where it should use 1-beta; Eq. (10) gets the beta right, so the two equations do not match. The perfect phase knowledge of both Bob and Eve (the LoP/DOA assumption) is another real limitation, and the paper gives no sensitivity analysis for phase errors. The lack of code, data, Monte-Carlo trial counts, or error bars makes the quantitative claims impossible to reproduce independently.\n\nAll of that said, the core idea is not silly. It is a plausible heuristic, and with a careful rewrite that either optimizes the magnitudes directly or explicitly treats the magnitude projection as a heuristic and backs it with reproducible simulations, it could become a useful short paper for the physical-layer security community. As it stands, the math and the simulations do not line up, and the reported advantage over equal-amplitude beamforming is not established.\n\nI would not desk-reject this out of hand, but I would send it to a referee with instructions to check the consistency between the optimization variables and the transmitted vectors, and to ask for code or detailed numerical tables. If the authors can fix the derivations or reframe the method honestly as a heuristic, the paper has a place. Right now it needs major revision before the results can be taken seriously.","headline":"The decoupling idea is a reasonable extension of the SPWT framework, but the paper's derivations describe complex vectors that are never actually transmitted, so the simulation claims are not supported by the math as written.","tokens_in":7237,"tokens_out":2670,"would_cite":false,"duration_ms":31723,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Decoupling phase and amplitude in beamforming yields two amplitude-only rules that beat equal-amplitude baselines in secrecy rate and bit-error rate at medium and high SNR.","keywords":["secure precise wireless transmission","precise jamming","phase alignment","amplitude beamforming","secrecy rate","bit error rate","artificial noise","directional modulation"],"falsifier":"Repeat the paper's 32-antenna, QPSK simulations with a controlled phase error added to the transmitter's steering phases — for example, uniformly distributed errors growing from a few degrees to tens of degrees. If the leakage and Max-RP secrecy-rate and bit-error-rate curves fall to or below the equal-amplitude curves for modest phase errors, the claim that these amplitude rules outperform EAB for precise jamming under phase alignment would be falsified.","tokens_in":6211,"feed_emoji":"📡","tokens_out":10580,"duration_ms":92809,"temperature":0.7,"pith_summary":"This paper asks whether secure wireless transmission can be improved by separating phase from amplitude in beamforming. It argues yes: once the phases of the message and artificial-noise vectors are aligned to the intended receiver and the eavesdropper, the remaining amplitude weights can be chosen by two simple criteria — maximum signal-to-leakage-noise ratio (leakage) and maximum received power under a null-space constraint (Max-RP). In simulations with a 32-antenna array, both proposed amplitude rules give better bit-error rate and secrecy rate than equal-amplitude beamforming at medium and high signal-to-noise ratios, and match it at low SNR. The practical stake is that precise jamming and precise communication can be achieved with closed-form eigenvector computations rather than joint phase-amplitude optimization.","feed_headline":"Two amplitude rules beat equal-amplitude jamming at mid and high SNR","feed_subtitle":"By fixing phase alignment first, simple eigenvector rules focus the message at Bob and noise at Eve, improving secrecy rate and BER.","key_machinery":"The load-bearing objects are the phase-aligned diagonal matrices $\\mathbf{P}$ and $\\mathbf{Q}$: $\\mathbf{P}$ carries the phases of the steering vector to Bob and $\\mathbf{Q}$ the phases to Eve, so the message vector is $\\mathbf{v}_{\\mathrm{CM}} = \\mathbf{P}\\mathbf{a}$ and the artificial-noise vector is $\\mathbf{v}_{\\mathrm{AN}} = \\mathbf{Q}\\mathbf{b}$. Once these phases are fixed, $\\mathbf{a}$ and $\\mathbf{b}$ are the only free amplitude weights. Leakage AB solves a maximum signal-to-leakage-noise-ratio problem whose solution is the dominant generalized eigenvector of a matrix pairing the desired channel with the eavesdropper channel plus noise. Max-RP AB uses the singular-value decomposition of the eavesdropper's phase-aligned channel to build a null-space projector, then picks the dominant eigenvector of the projected desired channel. These objects carry the argument because they turn a two-vector joint design into two one-vector amplitude designs with closed-form solutions.","core_discovery":"The central discovery is that the joint optimization of beamforming phase and amplitude can be decomposed without loss: for secure precise wireless transmission with precise jamming, phase alignment is mandatory, so the amplitude part alone remains free. The paper derives two amplitude rules. The leakage rule maximizes signal-to-leakage-plus-noise ratio, giving an eigenvector of a generalized matrix pencil. The Max-RP rule maximizes the desired receiver's power while forcing the signal onto the null space of the eavesdropper's channel, again via an eigenvector after a singular-value decomposition. Simulation results then show that both proposed rules outperform equal-amplitude beamforming in bit-error rate and secrecy rate at medium and high SNR, while all three behave similarly at low SNR. The paper also reports that all three rules create two distinct main peaks — CM around Bob and AN around Eve — which is its definition of precise communication and precise jamming.","pith_inferences":["A natural extension the paper leaves implicit is a sensitivity analysis: because the entire scheme rests on exact phase alignment, adding direction-of-arrival estimation error would be expected to shrink the claimed advantage over EAB, and the crossover error level is the quantity to measure.","Since the leakage rule contains an explicit noise-regularization term while Max-RP does not, the near-equality of the two schemes at the simulated SNRs suggests that term is negligible there; simulations at lower SNR or with unequal noise powers would reveal when the two rules genuinely differ.","The Max-RP null-space construction leaves $N-1$ degrees of freedom; in a scenario with several colluding eavesdroppers, the same construction could be applied to the joint null space of all their phase-aligned channels, at the cost of reduced array gain for the message."],"forward_implications":["A transmitter can approximate precise jamming without iterative joint optimization: fix the phases, then compute one eigenvector for the message and one for the noise.","The medium/high-SNR secrecy-rate gain means the proposed amplitude rules are most valuable precisely when the eavesdropper's channel is strong enough to threaten interception.","Because EAB, leakage, and Max-RP all form CM and AN peaks around Bob and Eve, even equal amplitudes already achieve the spatial focusing; the amplitude rules mainly sharpen those peaks.","At low SNR the three schemes coincide, which follows from the noise term dominating the amplitude-dependent terms in the optimization.","The near-identical performance of leakage and Max-RP suggests either criterion can be chosen on implementation grounds without sacrificing the stated gain."],"supporting_citations":[{"why":"Supplies the secure precise wireless transmission model with random subcarrier selection that the paper extends to precise jamming.","marker":"[8]"},{"why":"Introduces artificial-noise-aided secure transmission with random frequency diverse arrays, the secrecy idea behind precise jamming.","marker":"[7]"},{"why":"Defines frequency diverse arrays, the range-angle decoupling concept on which SPWT and PJ rest.","marker":"[9]"},{"why":"Provides the random-subcarrier-selection methods used in the RSS-PJ-AN system model.","marker":"[12]"},{"why":"Justifies the high-resolution direction-of-arrival estimation that supplies Bob's and Eve's locations to the phase alignment.","marker":"[14]"},{"why":"Defines the wiretap channel and the secrecy-rate metric used to evaluate the schemes.","marker":"[1]"},{"why":"Demonstrates the SVD-aided multi-beam directional modulation approach that motivates the null-space construction in Max-RP.","marker":"[13]"}],"fun_headline_variants":["Leakage and Max-RP amplitude rules beat equal beamforming at mid/high SNR","Phase alignment frees amplitude: two eigenvector rules boost secrecy rate","Eigenvector rules focus noise on Eve, signal on Bob, beating uniform power","Secure precise transmission: two amplitude rules outperform equal at high SNR"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire scheme assumes the transmitter knows the exact phase responses to both Bob and Eve — line-of-sight channels (printed as 'LoP') with high-resolution direction-of-arrival estimates — so that phase alignment is perfect; if these phases contain error, the focused peaks and the stated advantage over equal-amplitude beamforming degrade.","fun_headline_variants_meta":{"raw":{"variants":["Leakage and Max-RP amplitude rules beat equal beamforming at mid/high SNR","Phase alignment frees amplitude: two eigenvector rules boost secrecy rate","Eigenvector rules focus noise on Eve, signal on Bob, beating uniform power","Secure precise transmission: two amplitude rules outperform equal at high SNR"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000745,"raw_usage":{"total_tokens":3348,"prompt_tokens":1000,"completion_tokens":2348,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":616,"completion_tokens_details":{"reasoning_tokens":2266}},"tokens_in":616,"tokens_out":2348,"duration_ms":13947,"temperature":1.0,"reasoning_tokens":2266,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:18:11.734066+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the paper's 32-antenna, QPSK simulations with a controlled phase error added to the transmitter's steering phases — for example, uniformly distributed errors growing from a few degrees to tens of degrees. If the leakage and Max-RP secrecy-rate and bit-error-rate curves fall to or below the equal-amplitude curves for modest phase errors, the claim that these amplitude rules outperform EAB for precise jamming under phase alignment would be falsified.","supporting_citations":[{"cited_title":"Artifici al-noise- aided secure transmission with directional modulation bas ed on random frequency diverse arrays,","cited_arxiv_id":null,"evidence_quote":"Introduces artificial-noise-aided secure transmission with random frequency diverse arrays, the secrecy idea behind precise jamming."},{"cited_title":"F requency diverse array radars,","cited_arxiv_id":null,"evidence_quote":"Defines frequency diverse arrays, the range-angle decoupling concept on which SPWT and PJ rest."},{"cited_title":"Two practical random-subcarrier-selection methods for secure precise w ireless trans- missions,","cited_arxiv_id":null,"evidence_quote":"Provides the random-subcarrier-selection methods used in the RSS-PJ-AN system model."},{"cited_title":"SVD-aided multi- beam directional modulation scheme based on frequency dive rse array,","cited_arxiv_id":null,"evidence_quote":"Demonstrates the SVD-aided multi-beam directional modulation approach that motivates the null-space construction in Max-RP."}],"review_version":1}