Project supported by the National Natural Science Foundation of China (Grant No. 11747008) and the Natural Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2016GXNSFBA380227).
Project supported by the National Natural Science Foundation of China (Grant No. 11747008) and the Natural Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2016GXNSFBA380227).
† Corresponding author. E-mail:
Project supported by the National Natural Science Foundation of China (Grant No. 11747008) and the Natural Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2016GXNSFBA380227).
We investigate that no-knowledge measurement-based feedback control is utilized to obtain the estimation precision of the detection efficiency. We show that no-knowledge measurement is the optimal way to estimate the detection efficiency. The higher precision can be achieved for the lower or larger detection efficiency. It is found that no-knowledge feedback can be used to cancel decoherence. No-knowledge feedback with a high detection efficiency can perform well in estimating frequency and detection efficiency parameters simultaneously; simultaneous estimation is better than independent estimation given by the same probes.
Quantum metrology is a fundamental and important subject, which concerns the estimation of parameters under the constraints of quantum mechanics.[1–4] There are widespread applications such as in timing, healthcare, defence, navigation, and astronomy.[5–11] For the mean-square error criterion, the Cramér–Rao bound[12–14] is the most well-known analytic bound on the error. The estimation precision of the parameter is inversely proportional with quantum Fisher information (QFI).
The detection efficiency is a crucial physical quantity to judge the detector quality; it is becoming more and more important to improve the estimation accuracy of the detection efficiency.[15] Quantum feedback control[16–19] can be used to improve the detection efficiency. In Ref. [15], the use of the different measurement-based quantum feedback types to enhance the QFI about the detection efficiency of the detector has been investigated. In general, more signal, less noise from the detection can be used to better perform feedback control.[20,21] However, Szigeti et al.[22] showed significant and surprising results: performing a no-knowledge measurement (no signal, only noise) can be advantageous in canceling decoherence. It is due to that a system undergoing no-knowledge monitoring has reversible noise, which can be canceled by directly feeding back the measurement signal. For a perfect detection efficiency η = 1, no-knowledge feedback control can be used to completely cancel Markovian decoherence.
In this paper, we consider that the information of detection efficiency is encoded by a no-knowledge measurement-based quantum feedback. When one uses the optimal feedback operator to cancel decoherence, no-knowledge quantum feedback control is the optimal way to measure the detection efficiency. For the low or high detection efficiency, the precision can be dramatically high. Finally, we show that simultaneous estimation of the frequency and detection efficiency parameters has an advantage over estimating the noise and phase parameters individually, bringing us into the field of multi-parameter quantum metrology.[23,24] Also, no-knowledge quantum feedback can perform well in estimating the frequency with a high detection efficiency. Due to that the decoherence can be significantly canceled by the no-knowledge quantum feedback control with the high detection efficiency.
The rest of this article is arranged as follows. In Section
The famous Cramér–Rao bound[12,13] offers a very good parameter estimation under the constraints of quantum physics
If the probe state is pure,
For the classical multi-parameter Cramér–Rao bound:
Consider a two-dimensional system with Hamiltonian H interacting with a Markovian reservoir via the coupling operator L. The system density operator, ρ(t), evolves according to the master equation
For a homodyne measurement of the environment at angle θ, the conditional system dynamics are described by a stochastic master equation (SME)[28,29]
Then the control Hermitian can be written as Hfb = I(t)F where F is the feedback Hermitian operator, as shown in Fig.
For non-Hermitian operator L, the corresponding unconditional master equation is given by
For canceling decoherence, the optimal feedback operator is F = L in the case of Hermitian operator L.[22] In the case of non-Hermitian operator L, the optimal feedback operators are F± = L±.[22] We consider that the feedback operator is proportional to the coupling operator: F = λL (F± = λL±), where λ denotes a real number.
Firstly, we consider the Markoivian dephasing operator
We can see that the uncertainty of η can be 0 for η = 0. When λ = 1, the uncertainty of η can also be 0 for η = 1. Then, we can achieve the maximal QFI by taking the derivative of t and λ. However, it is difficult to calculate the exact analytical solution. We make an approximation for η < 1:
Then we can obtain the approximate optimal feedback constant λ = 1 and the interrogation time
We can find that the approximate solution in Eq. (
In general, the optimal feedback operator can be chosen to be the coupling operator (λ = 1).
Next, we consider the Markoivian dissipative operator
Without loss of generality, it is simply to choose ω = 0. Choosing
In general, simultaneous estimation of multi-parameter can perform better than estimating each parameter independently.[32] We consider simultaneously estimating frequency ω and detection efficiency η. The information of ω and η is encoded onto the evolved density matrix as shown in Eq. (
Utilizing Eqs. (
We numerically calculate the optimal precision of simultaneous and independent estimation with the same probes by choosing the optimal t and λ, as shown in Fig.
We have utilized the no-knowledge feedback control to estimate detection efficiency. The results show that when the feedback operator F is proportional to the coupling operator L, the no-knowledge measurement is the optimal way to encode the information of detection efficiency onto the probe state. By the exact numerical and approximate analytical calculation, we find that the high precision of detection efficiency can be obtained for low or large detection efficiency. Finally, we show that a simultaneous estimation frequency and detection efficiency with no-knowledge feedback control can perform better than independent estimation.
Whether no-knowledge measurement is the optimal way for any feedback operator will be our further investigation. In this article, we only consider the Markovian feedback. Non-Markovian phenomena are very common.[33,34] Hence, it is worth utilizing non-Markovian feedback to estimate the precision of detection efficiency.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] |