Time to Correct Match (TCM)

• Time to Correct Match (TCM) is a measure of how difficult it is to correctly associate signals or observations under delays, uncertainty, and structural constraints.
• In dual-phase TPC detectors, TCM is quantified as a product of event rate and delay window (M = R × T) to evaluate S1-S2 matching accuracy and mis-match rates.
• In applications like CRC-aided modulation and temporal graphs, TCM informs decoder search depth and recovery window design, optimizing performance and synchronization.

Searching arXiv for papers using the phrase “Time to Correct Match” and related “TCM” usages to ground the article in published work. Time to Correct Match (TCM) denotes, in the broadest sense, the latency, ambiguity, or iterative effort required to obtain the correct association between observations, signals, channels, or candidate outputs. The available literature suggests that the phrase is not a universally standardized technical term; instead, closely related papers operationalize it in domain-specific ways, including delayed signal association in dual-phase time projection chambers, CRC-aided list decoding of trellis coded modulation, inter-channel synchronization in particle detectors, and recovery-constrained matching in temporal graphs. At the same time, the acronym TCM is heavily overloaded in adjacent literatures, where it can mean transport channel matching, trellis coded modulation, or Temporal Cluster Matching, all of which are technically distinct from a time-to-correct-association notion (Radics et al., 2018, Wang et al., 2021, Abe et al., 11 Aug 2025, Mertzios et al., 2019, Zhang et al., 2022, Robinson et al., 2021).

1. Scope and operational meanings

A useful way to read TCM across the cited work is as an application-dependent measure of how difficult it is to arrive at the correct match under delay, uncertainty, or structural constraints. In some cases the quantity is explicit, such as the number of Viterbi searches needed before a CRC-valid codeword is found; in others it is implicit, such as the ambiguity induced by many candidate signals in a delay window or the number of iterations needed to synchronize detector channels.

Domain	Operationalization related to correct match	Representative paper
Dual-phase TPC detectors	Ambiguity of assigning delayed S2 to the correct preceding S1, quantified by $\mathcal{M}=\mathcal{R}\times\mathcal{T}$ and measured by mis-match fraction	(Radics et al., 2018)
CRC list decoding	Number of candidate paths examined until a correct CRC-passing codeword is found	(Wang et al., 2021)
Multi-channel detector calibration	Residual time difference between matched hits, reduced by iterative offset updates	(Abe et al., 11 Aug 2025)
Temporal graphs	Recovery window $\Delta$ constraining when a vertex may be matched again	(Mertzios et al., 2019)

A related timing notion also appears in surface-code decoding, where the minimum weight perfect matching implementation empirically requires only $O(n^2)$ average time per round for code distances ranging from 4 to 512, and is stated to be parallelizable to achieve $O(1)$ processing per round (Fowler et al., 2012). This is not presented as TCM terminology, but it is a direct study of the time required to complete correct matching in a fault-tolerant decoding pipeline.

2. Delayed signal association in dual-phase TPC detectors

In dual-phase TPC detectors, each physical event produces a fast primary (S1) light signal and a delayed secondary (S2) light signal. In large-scale detectors, the S2 emission can be delayed up to order of milliseconds, and this delay, combined with high signal rates, makes the matching of S1 and S2 challenging. The problem is formulated as a combinatorial optimization problem on a bipartite graph of eligible S1-S2 pairs, with rankings derived from a likelihood using event properties such as position, timing, and drift velocity. For single-scatter events the paper uses the Stable Marriage Problem; for multi-scatter events it uses the College Admission Problem, both solved by the Gale-Shapley algorithm (Radics et al., 2018).

Although the paper does not use the phrase “Time to Correct Match” explicitly, it operationalizes matching difficulty through the dimensionless parameter

$\mathcal{M}=\mathcal{R}\times\mathcal{T},$

where $\mathcal{R}$ is the event rate and $\mathcal{T}$ is the S2 delay time window. This parameter quantifies the expected number of S1s within the S2 delay window. The ranking likelihood is written as

$$L(S1\,|\,S2)\propto g(x_{s1}, t_{s1}\,|\,x_{s2}, t_{s2}, v_{\text{drift}}),$$

with $g$ described, for practical purposes, as a Gaussian in the residual between expected and observed S2 delay. Simulation studies show that for $\mathcal{M}<1$ the correct match is found nearly always, with mis-match $\Delta$0, while the mis-match rate increases exponentially as $\Delta$1 grows; at $\Delta$2 the error rate is $\Delta$3, increasing rapidly thereafter. Position smearing materially worsens performance: with $\Delta$4 position uncertainty, mis-match grows to $\Delta$5 at $\Delta$6. In the many-to-one case, average mis-match stays below $\Delta$7 for $\Delta$8 and grows to $\Delta$9 by $O(n^2)$0. In this setting, TCM is therefore best understood as a measure of association ambiguity driven by rate, delay, and ranking fidelity rather than as a single scalar latency.

3. Decoder search depth in CRC-aided trellis coded modulation

In short-blocklength communications, a closely related notion of TCM appears in CRC-aided Trellis Coded Modulation with Probabilistic Amplitude Shaping. The transmitter first applies a distribution matcher to generate amplitude symbols with the desired distribution, then appends CRC bits, and finally uses Ungerboeck’s TCM scheme, consisting of a $O(n^2)$1 systematic tail-biting convolutional code and a mapping function to channel signals. The receiver uses serial list Viterbi decoding (S-LVD), which enumerates candidate codewords in order of likelihood under the AWGN channel while incorporating the shaped prior (Wang et al., 2021).

Here, “time to correct match” is defined in the summary as the number of Viterbi searches, or equivalently the number of codeword candidates examined, until a correct CRC-passing codeword is found. The path metric used by the decoder is

$O(n^2)$2

After each candidate is generated, the decoder checks the CRC; if the candidate fails, the next-most-likely path is tested. The simulations use list size $O(n^2)$3. For the proposed CRC-TCM-PAS system with 87 input bits and 65–67 8-AM coded output symbols, the reported decoding performance under additive white Gaussian noise achieves the short-blocklength RCU bound with properly designed CRC and convolutional codes. The practical significance of TCM in this setting is therefore decoder latency: low average search depth translates directly into low-latency operation for short blocks.

4. Inter-channel synchronization in multi-channel particle detectors

A different but closely related formulation appears in time calibration for multi-channel particle detectors. Fixed per-channel offsets caused by cable lengths, electronics, and clock distribution limit overall timing resolution, and inter-channel calibration is required to synchronize hit times. In this context, the summary identifies TCM as the time difference between hits in two channels that result from a single physical event. The paper introduces an iterative calibration method based on Markov Chains that uses only correlated hit pairs and does not require an external reference time measurement (Abe et al., 11 Aug 2025).

Given a matching hit pair with measured times $O(n^2)$4 and expected time difference $O(n^2)$5, the discrepancy is

$O(n^2)$6

For channel $O(n^2)$7, the iteration updates are based on averaged discrepancies,

$O(n^2)$8

with matrix form

$O(n^2)$9

where $O(1)$0 is row-stochastic. The paper states that convergence follows from the Perron-Frobenius theorem, that synchronization is obtained up to an overall constant offset, and that the number of iterations to convergence is controllable with the damping factor $O(1)$1. The final offset precision is limited by intrinsic single-channel resolution, the number of matching hit pairs, and geometric or propagation uncertainties. In application, the method improves the SuperFGD single-fiber time resolution from 1.81 ns to 1.36 ns, and for top-bottom ToF panel pairs it improves resolution from 298 ps to 175 ps. In this detector-calibration setting, TCM is not a combinatorial ambiguity count but a residual synchronization error to be iteratively driven toward the intrinsic detector limit.

5. Recovery-constrained matching in temporal graphs

In temporal graph theory, the closest analogue to TCM is the recovery period that must elapse before a vertex can be matched again. A temporal graph is formalized as $O(1)$2, where $O(1)$3 assigns discrete time-labels to each edge, and a time-edge is a pair $O(1)$4. The Maximum Temporal Matching problem seeks the largest subset of time-edges such that no vertex is matched more than once within any time window of $O(1)$5 consecutive time slots. Equivalently, two chosen time-edges must either be edge-disjoint or have time-labels at least $O(1)$6 apart (Mertzios et al., 2019).

The summary explicitly links this recovery period $O(1)$7 to a TCM-like notion in dynamic networks. Computationally, the problem is difficult: it is NP-complete for $O(1)$8 even when $O(1)$9 and each edge is active in exactly one time slot, and it is APX-hard. The paper defines a $\mathcal{M}=\mathcal{R}\times\mathcal{T},$0-temporal line graph in which vertices are time-edges and conflicts become graph edges, so that a $\mathcal{M}=\mathcal{R}\times\mathcal{T},$1-temporal matching corresponds to an independent set. To cope with hardness, it gives a polynomial-time approximation algorithm with ratio

$\mathcal{M}=\mathcal{R}\times\mathcal{T},$2

which is $\mathcal{M}=\mathcal{R}\times\mathcal{T},$3 for $\mathcal{M}=\mathcal{R}\times\mathcal{T},$4 and approaches $\mathcal{M}=\mathcal{R}\times\mathcal{T},$5 for large $\mathcal{M}=\mathcal{R}\times\mathcal{T},$6. It also provides fixed-parameter algorithms with runtimes $\mathcal{M}=\mathcal{R}\times\mathcal{T},$7 when parameterized by solution size $\mathcal{M}=\mathcal{R}\times\mathcal{T},$8, and $\mathcal{M}=\mathcal{R}\times\mathcal{T},$9 when parameterized by $\mathcal{R}$0, where $\mathcal{R}$1 is the maximum static matching size of the underlying graph. In this literature, TCM is best read as a refractory window embedded directly into the feasibility constraints of temporal matching.

6. Exactness, near misses, and acronym disambiguation

A related evaluation problem arises in temporal question answering, where answers are numeric but systems are often scored with Exact Match (EM). The paper “Time to Revisit Exact Match” argues that EM is unable to distinguish small from large errors, and evaluates models instead with sMAPE and MASE on the numeric-only TempAnswerQA benchmark of 3,434 QA pairs. The reported findings are that error size and EM are decoupled, that models with low EM can still have low sMAPE, that MASE reshuffles model rankings compared to EM, and that the most frequent error is to deviate by only $\mathcal{R}$2 from the ground truth (Abbood et al., 20 Sep 2025). This suggests that whenever TCM is interpreted as progress toward a correct association or value, binary “correct/incorrect” criteria can obscure informative near misses.

The acronym TCM also requires explicit disambiguation. In robust JPEG steganography, transport channel matching is an iterative recompression procedure that yields a channel-matched image whose DCT coefficients are invariant, or nearly invariant, to further recompression; the cited work then augments adaptive embedding with a dynamically updated robustness cost and integrates it with STC embedding (Zhang et al., 2022). In remote sensing, Temporal Cluster Matching is a model for change detection in structures from satellite imagery, based on comparing cluster distributions inside a footprint and its neighborhood and detecting the earliest time at which their KL-divergence exceeds a threshold (Robinson et al., 2021). In communications, TCM commonly abbreviates trellis coded modulation rather than any measure of time to a correct match (Wang et al., 2021). A common misconception is therefore to treat TCM as a single canonical object across fields. The literature instead supports a narrower conclusion: “time to correct match” is a useful cross-domain interpretive label for latency-to-correct-association phenomena, but the acronym TCM is semantically overloaded and must be resolved from context.