1: %#!platex journal.tex
2: %%
3: %% analysis_model.tex %%
4: %%
5: %% Last modified: 05/03/25 17:55:07 on dell148
6: %%
7:
8: \subsection{Analysis model for the decoherence}
9: \label{sec:analysis_method}
10:
11: The probability that
12: the decoherence error occurs is considered to
13: increase in proportion to the product
14: of execution time and the number of qubits.
15: As stated in Section~\ref{sec:counting},
16: the quantum counting is performed in five stages:
17: (\ref{enu:counting_1}) preparing an initial state,
18: (\ref{enu:counting_2}) applying the Hadamard transformation,
19: (\ref{enu:counting_3}) applying controlled-$G$ operations,
20: (\ref{enu:counting_4}) applying the inverse quantum Fourier transform,
21: and (\ref{enu:counting_5}) measurements.
22: The time needed for each stage is as follows:
23: The stage~(\ref{enu:counting_1}) and (\ref{enu:counting_5}) are considered
24: to be one step.
25: Application of the Hadamard transformations
26: on the stage (\ref{enu:counting_2})
27: takes $O(1)$ and
28: application of the inverse quantum Fourier transform
29: on the stage (\ref{enu:counting_4})
30: takes at most $O(p^2)$.
31: In contrast, application of controlled-$G$ operations
32: on the stage~(\ref{enu:counting_3}) takes $2^p-1$, which is exponential to $p$.
33: It follows that the decoherence error happens on the stage~(\ref{enu:counting_3})
34: with exponentially higher probability
35: than on other stages.
36: On our analyses, we restrict
37: the position of the error only on the stage~(\ref{enu:counting_3}).
38:
39: %%% エラーの数(1個) -> 仮定2
40: %We then notice the number of the decoherence errors.
41: %Quantum algorithms are greatly affected
42: %by the decoherence,
43: %and we cannot obtain the desired output
44: %if the errors occur more than several times.
45: Suppose that the error rate $d$ is small enough $(d \ll 1)$,
46: then influences of the decoherence are approximated
47: by the first order term of $d$.
48: %In order to analyze the most serious influences
49: %of the decoherence, that is the first order term of the error,
50: %it is sufficient to consider the following case:
51: This means that the depolarizing channel is applied only once.
52: %
53: %\begin{assumption}[the number of decoherence errors]
54: % \label{asmp:error_once}
55: % The decoherence errors occur {\itshape only once}
56: % on the quantum counting.
57: %\end{assumption}
58: %
59: Since the quantum counting has two registers,
60: the register where the depolarizing channel is applied
61: is either of the first register or the second one.
62: %we need only to consider two cases that
63: %the error occurs on either of the first register or the second one.
64: %Finally, we explain error models for our analyses.
65: %
66: %%% エラーモデル -> 仮定3
67: %\begin{assumption}[decoherence model]
68: % On the first register,
69: % we assume a decoherence model as the depolarizing channel.
70: % On the second register,
71: % we have no assumption of type of the decoherence errors.
72: % %but assume that the decoherence affects only the second register.
73: %\end{assumption}
74: %
75: %In the case of the first register,
76: %we use the depolarizing channel
77: %often used for analyzing
78: %quantum error correction codes.
79: %Since $\sigma_x, \sigma_y$, and $\sigma_z$ errors
80: %occur with the same probability through the depolarizing channel,
81: %%By Definition~\ref{def:depolarizing}
82: %%and Assumption~\ref{asmp:error_once},
83: %we deal with the decoherence
84: %such that we take the average of influences
85: %in three error cases
86: %that each error occurs once on the first register.
87: %On the other hand,
88: %in the case of the second register,
89: %we analyze the decoherence
90: %by no assumption of error models
91: %since we cannot analyze it in detail
92: %even if we assume some error model
93: %because of quantum oracle,
94: %noted later in Section~\ref{sec:analysis_second}.
95:
96: Under these conditions,
97: we calculate probability distributions on the quantum counting,
98: and give some properties of the decoherence on two registers.
99: Moreover, we compare influences of the decoherence
100: on the ascending-order and the descending-order implementations.
101: %defined in Section~\ref{sec:two_implementations}.
102: %We analyze the decoherence on the quantum counting
103: %related to not only the first and the second register
104: %but also the ascending-order and the descending-order
105: %implementation,
106: %and compare influences of the decoherence
107: %between two implementations.
108: