{"id":"bf52c094-fa72-4357-9f8b-767b7fe76838","arxiv_id":"1908.05798","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A construction that balances polar codewords with Knuth's algorithm gives flicker-free VLC transmission at 50% dimming with logarithmic redundancy and BER gains over RLL-based schemes.","lead":"Visible light communication sends data through LED brightness changes, but flicker and dimming force extra coding overhead. This paper combines polar error-correcting codes with Knuth's balancing algorithm to create LED-friendly codes with less overhead and better error correction than standard methods.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Flicker-free claim is falsified by the all-zero information word, which after Knuth balancing produces a run of N/2 zeros.","rationale":"The reader's weakest assumption identified the statistical, rather than deterministic, nature of the run-length analysis. This stress-test confirms and sharpens that concern with an explicit counterexample. The proposed scheme's encoding path is deterministic: an all-zero message maps to an all-zero polar codeword, and Knuth balancing of 0^N necessarily yields 1^{N/2}0^{N/2}. This is not a rare event under a non-uniform adversarial input; it is a legal codeword in any FEC system. At the paper's own lower clock-rate scenario, this run exceeds the MFTP and the eye-safe switching frequency, so flicker can occur. The empirical study in Section III-C, based on random information words, cannot exclude this deterministic worst case, and extrapolating a 10^-7 failure rate from a single observed event at run length 28 is not a substitute for a tail bound. The FER formula and BER simulations are internally plausible and match simulation, but they address error-correction performance, not the run-length constraint. Because the central claim of flicker-free coding is falsified for a valid input, the paper's main advertised advantage over RLL schemes fails. A revised scheme could add an explicit run-length limiter or a Knuth variant with bounded maximal run, and then the performance analysis could be re-evaluated; as written, the central claim does not hold.","tokens_in":8947,"tokens_out":7213,"duration_ms":78541,"concrete_test":"Run the proposed encoder exactly as specified in Section III-A with u = 0 for PC(2048,1638), compute the Knuth-balanced output x' and the transmitted sequence s = x'p'p'', and measure the maximum run length. If the maximum run equals 1024 (or exceeds 28), Section III-C's empirical maximum-run claim is falsified and the flicker-free guarantee fails for a legal information word. This check requires only one encoder call and directly settles whether the all-zero counterexample applies to the implemented scheme.","verdict_should_be":"REJECT","load_bearing_attack":"The paper's central claim that the proposed scheme generates flicker-free codes is not merely unproven; it is false for a valid input. For any polar code PC(N,K), the all-zero information word u=0 encodes to the all-zero codeword x=0^N. Knuth's balancing algorithm then flips the first N/2 bits, producing x' = 1^{N/2} 0^{N/2}, which contains a run of length N/2. For the PC(2048,1638) used in Section III-C, this run length is 1024, not the empirically reported maximum of 28. Section III-C only tests 10,000 uniformly random information words and extrapolates a failure rate of about 10^-7 at run length 28; it does not bound the worst case. Under the paper's own 'lowest optical clock rate of 200 kHz' assumption, a 1024-bit run lasts 5.12 ms, exceeding the 5 ms maximum flickering time period, and the corresponding switching frequency is about 195 Hz, below the 200 Hz eye-safe threshold. Since FEC codes must handle all information words, including all-zero data, the deterministic flicker-free guarantee is invalid. The BER and complexity comparisons in Section III-D and Table I may still be valid, but the headline advantage over RLL codes, which guarantee run-length constraints, is not established by the proposed scheme.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a concatenated coding scheme for visible light communication (VLC): an outer polar code, followed by Knuth's balancing algorithm, with the balancing prefix protected by a second polar code and a Manchester-coded complement appended. The authors claim that the transmitted sequence is flicker-free without lookup tables, with logarithmic redundancy, lower decoding complexity than RLL-based schemes, and BER gains of up to 1.8 dB over rate-matched polar+1b2b and polar+4b6b baselines. The BER and FER claims are supported by simulations and a Gaussian-approximation formula, and the complexity comparison is quantified. The flicker-free claim is supported only by an empirical run-length study over 10,000 uniformly random information words, with no worst-case bound.","tokens_in":9226,"tokens_out":7094,"duration_ms":64227,"significance":"The combination of polar codes with Knuth balancing is a natural idea, and the paper's rate-matched BER simulations (Fig. 7), the matching theoretical FER formula (4) and simulation (Fig. 6), and the complexity table (Table I) are credible and clearly presented. If flicker-free operation were guaranteed, the scheme would indeed offer an attractive alternative to RLL-based VLC codes in terms of redundancy and decoder complexity. However, the central advertised property, flicker-free communication, is not actually guaranteed by the construction. The run-length analysis of Section III-C is statistical and contains no deterministic bound, and a concrete valid input, the all-zero information word, produces a run of N/2 consecutive ones after balancing, which violates the paper's own VLC flicker criteria for the parameters used in the paper. Because the flicker-free claim is the main differentiator relative to RLL-based methods, this is a load-bearing failure.","major_comments":[{"comment":"The claim that the scheme generates flicker-free codes is false for a valid input. For any polar code PC(N,K), the all-zero information word u = 0^N encodes to the all-zero codeword x = 0^N. Knuth balancing then flips the first N/2 bits to achieve zero disparity, producing x' = 1^{N/2} 0^{N/2}, which contains a run of length N/2. For PC(2048,1638), the configuration studied in Section III-C, this run length is 1024, not the empirically reported maximum of 28. Using the paper's stated lowest optical clock rate of 200 kHz, a 1024-bit run corresponds to a period of 5.12 ms and a switching frequency of about 195 Hz, which is below the 200 Hz eye-safe threshold and above the 5 ms MFTP. The empirical study of Section III-C used only uniformly random information words and therefore missed this deterministic counterexample. Since FEC codes must operate on all information words, the abstract's and Section III-D2's statements that the proposed scheme generates flicker-free codes are not valid.","section":"III-C and III-D2"},{"comment":"The RLL failure rate analysis is purely empirical and provides no worst-case guarantee. The paper counts run lengths over 10,000 trials and then extrapolates the failure rate at l = 28 to about 10^-7 without any analytic tail bound. VLC flicker mitigation requires a deterministic constraint on the maximum run length or on the maximum flickering time period, not a probabilistic statement. The counterexample of the all-zero word shows that a deterministic bound cannot be derived for the current scheme as is; the authors would need to add an explicit run-length limiter or modify the balancing process, which would change the redundancy and complexity claims of the paper. As written, the conclusion that the scheme 'guarantee[s] a flicker-free communication' is unsupported.","section":"III-C"}],"minor_comments":[{"comment":"The example in (2) uses '101111' while the surrounding text refers to '1011111', and the phrase 'balanced state occurs at index 4 (100)' is unclear because the first four bits form 0100 after flipping. Please reconcile the example and explicitly define that the first e bits are the flipped segment.","section":"II-D"},{"comment":"The caption contains a typo: 'LPDC' should read 'LDPC'.","section":"Fig. 2 caption"},{"comment":"The symbol p is used both for the length of the Knuth prefix (p = log2 N) and for the encoded prefix (e.g., PC(p',p) and p''). Distinct notation for the prefix length and the encoded prefix would reduce ambiguity.","section":"III-A"},{"comment":"The sentence 'About 90% in each generated codeword have a run-length of l < 8' is imprecise; it should state whether this refers to 90% of all runs, or 90% of the transmitted bits, or another quantity.","section":"III-C"}],"recommendation":"reject","confidential_remarks":"The all-zero codeword counterexample is decisive: the paper's headline claim of generating flicker-free codes is false for a valid input under the paper's own parameters. The BER and complexity results may still be useful, but the central contribution is invalidated. A revision that adds a deterministic run-length limiting mechanism would essentially constitute a new scheme, so I do not see a path to acceptance within the current manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the paper is a clean engineering combination of Knuth's balancing algorithm and polar codes for VLC, with credible rate-matched BER simulations and a neat FER formula. But the central \"flicker-free\" claim is not just unproven—it is false for a valid input. The all-zero information word polar-encodes to all zeros; Knuth's balancing then flips the first N/2 bits, producing 1^{N/2}0^{N/2}, a run of length 1024 for their PC(2048,1638). At a 200 kHz optical clock that lasts 5.12 ms, above the 5 ms MFTP, and the switching frequency is about 195 Hz, below the eye-safe 200 Hz. Their empirical run-length study (10,000 random frames, max run 28) simply misses this deterministic worst case. The headline advantage over RLL codes—deterministic run-length guarantees—is therefore not established. This is a load-bearing flaw, not a nit. The fix is straightforward: add an explicit run-length limiter (bit-stuffing, block inversion, etc.) or honestly reframe the claim as \"balanced with typically short runs.\"\n\nWhat the paper does well: the redundancy grows logarithmically instead of linearly as in RLL codes; the theoretical FER in (4) matches simulation in Fig. 6; the BER gains over 1b2b and 4b6b baselines (up to 1.8 dB at R=0.44) are credible; the complexity table is useful. The construction itself is new relative to the cited polar-VLC literature (compensation symbols, modified kernels, RLL).\n\nSoft spots beyond the worst case: the BER comparison omits the closest prior scheme—Fang et al.'s polar code with compensation symbols [8]—which is the natural baseline for a flicker-free FEC at 50% dimming. Also, the run-length analysis considers only randomly generated frames; a FEC claim must hold for all information words, including all-zero data.\n\nThis paper is for VLC system designers interested in balanced FEC constructions, and also as a cautionary lesson on statistical versus deterministic constraints. It deserves a serious referee because the core tools are sound and the flaw is fixable. I'd send it out, with a clear request to address worst-case run-length behavior and to compare against [8].","headline":"Clever combination of Knuth balancing with polar codes, but the flicker-free claim fails on the all-zero codeword, which yields a run of N/2 bits.","tokens_in":9744,"tokens_out":2961,"would_cite":false,"duration_ms":29010,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"By combining polar codes with Knuth's balancing algorithm, this paper constructs flicker-free forward error correction codes for visible light communication that need no lookup tables and report BER gains up to 1.8 dB over regular schemes…","keywords":["visible light communication","flicker-free codes","Knuth balancing algorithm","polar codes","run-length limited codes","forward error correction","successive cancellation decoding","dimming control"],"falsifier":"Generate more than ten million transmitted frames of PC(2048,1638) under the proposed scheme and record the longest run of identical bits; a run length clearly above the paper's reported maximum of 28, or a measured failure rate at run length 28 above about $10^{-7}$, would show that the empirical run-length tail is heavier than the extrapolation assumes.","tokens_in":8757,"feed_emoji":"💡","tokens_out":12945,"duration_ms":116741,"temperature":0.7,"pith_summary":"This paper proposes a forward error correction scheme for visible light communication that prevents LED flicker without the lookup tables and rate penalties of run-length limited codes. The transmitter balances each polar-coded block with Knuth's balancing algorithm, appends a separately protected prefix that records which bits were flipped, and sends that prefix in a self-balancing form. The authors claim that at 50% dimming the resulting stream is flicker-free, uses only logarithmic redundancy, and supports higher transmission rates than the low-rate RLL codes. Their headline simulation result is a 1.8 dB and 0.9 dB gain in bit error rate at $10^{-6}$ over the regular schemes at rates 0.44 and 0.23.","feed_headline":"Knuth balancing makes polar codes flicker-free for LED links","feed_subtitle":"The scheme removes lookup tables, cuts redundancy to log scale, and reports up to 1.8 dB BER gain at 10^-6.","key_machinery":"Knuth's balancing algorithm: given a binary word x mapped to bipolar values, flip the first e bits at the unique index where the running digital sum crosses zero, producing a balanced word x' with equal numbers of +1 and -1, then append a binary prefix p that encodes e. This is what enforces the 50% duty cycle that suppresses flicker, and the raw balancing metadata costs only log2 N bits, which keeps rates high compared with linear-redundancy RLL codes. The paper adds a second protective layer: the prefix p is itself polar-encoded into p', and p'' is the bitwise complement of p', so the whole transmitted frame remains balanced while the receiver obtains soft information about the prefix from the difference p'_i - p''_i.","core_discovery":"The central claim is that balancing a polar codeword after encoding can enforce the 50% duty cycle that prevents visible flicker, without the memory and rate cost of RLL codes. In the proposed scheme, the balancing index prefix is itself polar-encoded and balanced by appending its bitwise complement, so every transmitted frame s = x'·p'·p'' has equal numbers of ones and zeros, and the receiver can decode the prefix with soft decisions before undoing the outer balancing. The paper reports redundancy that grows logarithmically with block length, transmission rates up to 0.75 at 50% dimming, and BER gains over 1b2b and 4b6b based schemes, with the strongest stated gains being 1.8 dB and 0.9 dB over the regular schemes at $10^{-6}$ for rates 0.44 and 0.23. On its own terms, the discovery is that the balancing constraint and the FEC can be separated cleanly, with the balancing metadata treated as data and protected by its own polar code.","pith_inferences":["The flicker-free guarantee is only as strong as the run-length tail: the paper extrapolates a failure rate near 10^-7 at run length 28 from 10,000 trials, so deployment would require an analytic bound or an explicit run-length limiter.","Knuth balancing fixes the ones density at exactly 50%, so dimming levels other than 50% would need a different balancing target or compensation symbols, which the paper leaves to future work.","The complement-pair trick used to protect the prefix is effectively a Manchester code for metadata; the same idea could protect frame headers or control fields in other constrained channels.","A natural extension is to add a bounded run-length constraint after balancing and measure the extra redundancy it costs, which would convert the empirical flicker-free claim into a guarantee."],"forward_implications":["VLC transmitters can drop memory-hungry RLL lookup tables, since balancing is computed arithmetically from each codeword.","Redundancy scales as O(log N) instead of O(N), so longer blocks make the scheme increasingly rate-efficient relative to 1b2b, 4b6b, and 8b10b codes.","Because the construction works with any FEC code, the same prefix-protected balancing can be applied to LDPC or turbo codes.","The concatenated FER expression lets a designer choose the inner prefix code size to trade prefix protection against rate loss.","For the simulated configurations, measured run lengths stay far below the 5 ms human-eye threshold, supporting the flicker-free label at those block lengths."],"supporting_citations":[{"why":"Supplies the balancing algorithm that turns each polar codeword into a zero-disparity word with a short prefix.","marker":"[12]"},{"why":"Defines the earlier polar-code flicker-free construction and the 10,000-test run-length methodology that this paper compares against and repeats.","marker":"[8]"},{"why":"Provides the IEEE 802.15.7 VLC standard, the RLL codes used as baselines, and the flicker safety threshold.","marker":"[4]"},{"why":"Introduces polar codes and successive cancellation decoding, which form the outer FEC and the decoder used throughout.","marker":"[17]"},{"why":"Gives the Gaussian-approximation FER prediction that the paper extends to the concatenated prefix-plus-information FER.","marker":"[20]"},{"why":"Provides the shortening technique that lets polar codes match the non-power-of-two lengths needed for rate comparisons.","marker":"[21]"}],"fun_headline_variants":["Polar codes + Knuth balancing kill VLC flicker without lookup tables","Balanced polar codes boost VLC rates and cut BER by 1.8 dB","Knuth's trick makes polar codes flicker-free for LED links","No lookup tables: Knuth-balanced polar codes for flicker-free VLC","Higher-rate flicker-free VLC with Knuth-balanced polar codes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Flicker-freedom is asserted from observed run lengths in 10,000 trials rather than from a proven bound on the longest run; if any transmitted frame contains a run long enough to push the switching period toward or beyond the eye-safety limit, the guarantee breaks, and the paper does not bound that worst case.","fun_headline_variants_meta":{"raw":{"variants":["Polar codes + Knuth balancing kill VLC flicker without lookup tables","Balanced polar codes boost VLC rates and cut BER by 1.8 dB","Knuth's trick makes polar codes flicker-free for LED links","No lookup tables: Knuth-balanced polar codes for flicker-free VLC","Higher-rate flicker-free VLC with Knuth-balanced polar codes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1505,"prompt_tokens":964,"completion_tokens":541,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":580,"completion_tokens_details":{"reasoning_tokens":442}},"tokens_in":580,"tokens_out":541,"duration_ms":4925,"temperature":1.0,"reasoning_tokens":442,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:05:03.529445+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Generate more than ten million transmitted frames of PC(2048,1638) under the proposed scheme and record the longest run of identical bits; a run length clearly above the paper's reported maximum of 28, or a measured failure rate at run length 28 above about $10^{-7}$, would show that the empirical run-length tail is heavier than the extrapolation assumes.","supporting_citations":[{"cited_title":"Efficient balanced codes,","cited_arxiv_id":null,"evidence_quote":"Supplies the balancing algorithm that turns each polar codeword into a zero-disparity word with a short prefix."},{"cited_title":"An efficient flicker-free FEC coding scheme for dimmable visible light communication based on polar codes,","cited_arxiv_id":null,"evidence_quote":"Defines the earlier polar-code flicker-free construction and the 10,000-test run-length methodology that this paper compares against and repeats."},{"cited_title":"IEEE standard for local and metropolitan area networks–part 15.7: Short-range wireless optical communication using visible light,","cited_arxiv_id":null,"evidence_quote":"Provides the IEEE 802.15.7 VLC standard, the RLL codes used as baselines, and the flicker safety threshold."},{"cited_title":"Channel polarization: a method for constructing capacity- achieving codes for symmetric binary-input memoryless channels,","cited_arxiv_id":null,"evidence_quote":"Introduces polar codes and successive cancellation decoding, which form the outer FEC and the decoder used throughout."},{"cited_title":"Construction and block error rate analysis of polar codes over awgn channel based on gaussian approximation,","cited_arxiv_id":null,"evidence_quote":"Gives the Gaussian-approximation FER prediction that the paper extends to the concatenated prefix-plus-information FER."},{"cited_title":"A novel puncturing scheme for polar codes,","cited_arxiv_id":null,"evidence_quote":"Provides the shortening technique that lets polar codes match the non-power-of-two lengths needed for rate comparisons."}],"review_version":1}