Quantum metrology with unitary parametrization processes

# Quantum metrology with unitary parametrization processes

Jing Liu Zhejiang Institute of Modern Physics, Department of Physics, Zhejiang University, Hangzhou 310027, China    Xiao-Xing Jing Zhejiang Institute of Modern Physics, Department of Physics, Zhejiang University, Hangzhou 310027, China    Xiaoguang Wang Zhejiang Institute of Modern Physics, Department of Physics, Zhejiang University, Hangzhou 310027, China Synergetic Innovation Center of Quantum Information and Quantum Physics, University of Science and Technology of China, Hefei, Anhui 230026, China
###### Abstract

Quantum Fisher information is a central quantity in quantum metrology. We discuss an alternative representation of quantum Fisher information for unitary parametrization processes. In this representation, all information of parametrization transformation, i.e., the entire dynamical information, is totally involved in a Hermitian operator . Utilizing this representation, quantum Fisher information is only determined by and the initial state. Furthermore, can be expressed in an expanded form. The highlights of this form is that it can bring great convenience during the calculation for the Hamiltonians owning recursive commutations with their partial derivative. We apply this representation in a collective spin system and show the specific expression of . For a simple case, a spin-half system, the quantum Fisher information is given and the optimal states to access maximum quantum Fisher information are found. Moreover, for an exponential form initial state, an analytical expression of quantum Fisher information by operator is provided. The multiparameter quantum metrology is also considered and discussed utilizing this representation.

###### pacs:
03.67.-a, 03.65.Ta, 06.20.-f.

How to precisely measure the values of physical quantities, such as the phases of light in interferometers, magnetic strength, gravity and so on, is always an important topic in physics. Obtaining high-precision values of these quantities will not only bring an obvious advantage in applied sciences, including the atomic clocks, physical geography, civil navigation and even military industry, but also accelerate the development of fundamental theories. One vivid example is the search for gravitational waves. Quantum metrology is such a field attempting to find optimal methods to offer highest precision of a parameter that under estimation. In recently decades, many protocols and strategies have been proposed and realized to improve the precisions of various parameters 1 (); 2 (); 3 (); 4 (); 5 (); 6 (); 7 (); 8 (); 9 (); 10 (); 11 (); 12 (); 13 (); 14 (); Yao (); Xiao (); 15 (); 16 (); Escher (); Vidrighin (); Braun (); Giovannetti06 (); Boixo (); Taddei (). Some of them can even approach to the Heisenberg limit, a limit given by the quantum mechanics, showing the power of quantum metrology.

Quantum Fisher information is important in quantum metrology because it depicts the theoretical lowest bound of the parameter’s variance according to Cramér-Rao inequality Helstrom (); Holevo (). The quantum Fisher information for parameter is defined as , where is a density matrix dependent on and is the symmetric logarithmic derivative (SLD) operator and determined by the equation . For a multiparameter system, the counterpart of quantum Fisher information is called quantum Fisher information matrix , of which the element is defined as , where , are the SLD operators for parameters and , respectively.

Recently, it has been found Pang () that quantum Fisher information can be expressed in an alternative representation, that all information of parametrization process in quantum Fisher information is involved in a Hermitian operator . This operator characterizes the dynamical property of the parametrization process, and totally independent of the selection of initial states. Utilizing this representation, the quantum Fisher information is only determined by and the initial state.

In this report, we give a general expression of quantum Fisher information and quantum Fisher information matrix utilizing operator. For a unitary parametrization process, can be expressed in an expanded form. This form is particularly useful when the Hamiltonian owns a recursive commutation relation with its derivative on parameter estimation. We calculate the specific expression of in a collective spin system, and provide an analytical expression of quantum Fisher information in a spin-half system for any initial state. Based on this expression, all optimal states to access maximum quantum Fisher information are found in this system. Furthermore, considering this spin-half system as a multiparameter system, the quantum Fisher information matrix, can be easily obtained by the known form of in single parameter estimations. On the other hand, inspired by a recent work Jiang14 (), for an exponential form initial state, we provide an analytical expression of quantum Fisher information using operator. A demonstration with a spin thermal initial state is given in this scenario. The maximum quantum Fisher information and the optimal condition are also discussed.

Results
Quantum Fisher information with operator.
For a general unitary parametrization transformation, the parametrized state is expressed by , where is a state independent of . In this paper, since the parameter is only brought by , not the initial state , we use instead of for short. Denote the spectral decomposition of as , where and are the th eigenvalue and eigenstate of and is the dimension of the support of . It is easy to see that and are the corresponding eigenvalue and eigenstate of , respectively. The quantum Fisher information for can then be expressed by Commuliu (); PhysicaAliu ()

 F=M∑i=14pi⟨Δ2H⟩i−∑i≠j8pipjpi+pj|⟨ψi|H|ψj⟩|2, (1)

where

 H:=i(∂αU†)U (2)

is a Hermitian operator since the equality . Meanwhile,

 ⟨Δ2H⟩i=⟨ψi|H2|ψi⟩−⟨ψi|H|ψi⟩2 (3)

is the variance of on the th eigenstate of . When commutes with , can be explained as the generator of the parametrization transformation Pang (). The expression (1) of quantum Fisher information is not just a formalized representation. The operator is only determined by the parametrization process, that is the dynamics of the system or the device. For a known dynamical process of a parameter, i.e., known system’s Hamiltonian, is a settled operator and can be obtained in principle. In this representation, the calculation of quantum Fisher information is separated into two parts: the diagonalization of initial state and calculation of . For a general 2-dimensional state, the quantum Fisher information reduces to

 Fqubit=4(2Trρ2−1)⟨Δ2H⟩1(2). (4)

The subscript of the variance can be chosen as 1 or 2 as any Hermitian operator’s variances on two orthonormal states are equivalent in 2-dimentional Hilbert space. For a pure state, the quantum Fisher information can be easily obtain from Eq. (4) with taking the purity and the variance on that pure state, i.e., Pang ()

 Fpure=4⟨Δ2H⟩in. (5)

Namely, the quantum Fisher information is proportional to the variance of on the initial state. In this scenario, denote the initial state , the quantum Fisher information can be rewritten into , with the effective SLD operator

 Leff=i2[H,|ψ0⟩⟨ψ0|]. (6)

For a well applied form of parametrization transformation  Pang (), where has been set as 1 in Planck unit, and being aware of the equation

can then be expressed by

 H=−∫t0eisHα(∂αHα)e−isHαds. (8)

Defining a superoperator as , can be written in an expanded form

 H=i∞∑n=0fnH×nα(∂αHα), (9)

where the coefficient

 fn=(it)n+1(n+1)!. (10)

In many real problems, the recursive commutations in Eq. (9) can either repeat or terminate Jiang14 (), indicating an analytical expression of . Thus, this representation of quantum Fisher information would be very useful in these problems. For the simplest case that , all terms vanish but the first one, then . When , with a constant matrix or proportional to , only the first and second terms remain. In this case, reduces to . A more interesting case is that , with a nonzero constant number, then can be written in the form

 H=ic[exp(itc)−1]∂αHα. (11)

In the following we give an example to exhibit Eq. (9). Consider the interaction Hamiltonian of a collective spin system in a magnetic field

 Hθ=B(cosθJx+sinθJz)=BJn0, (12)

where with and . is the amplitude of the external magnetic field and is the angle between the field and the collective spin. Here for with the Pauli matrix for th spin. Taking as the parameter under estimation, can be expressed by

 H=2∣∣∣sin(Bt2)∣∣∣Jn1, (13)

where with the vector

where is the sign function and is normalized.

The operator for Hamiltonian (12) may be also available to be solved using the procedure in Ref. Pang (), in the (2j+1)-dimensional eigenspace of (j is the total spin). In principle, the eigenstates of can be found by rotating the Dicke state into the same direction of . However, even one can analytically obtain all the eigenvalues and eigenvectors, it still requires a large amount of calculations to obtain , especially when the spin numbers are tremendous. Comparably, utilizing Eq. (9), it only takes a few steps of calculation, which can be found in the method. This is a major advantage of the expanded form of .

Utilizing Eq. (13), one can immediately obtain the form of for a spin-half system

 Hqubit=∣∣∣sin(Bt2)∣∣∣n1⋅σ, (14)

with , which was also discussed in the Hamiltonian eigenbasis in Ref. Pang (). For any 2-dimensional state, based on Eq. (4), the quantum Fisher information can be expressed by

 Fθ=4sin2(Bt2)|rin|2[1−(n1⋅re)2], (15)

where is the Bloch vector of the initial state and is the Bloch vector of any eigenstate of . For pure states, there is and . Since the Bloch vector of a 2-dimensional state satisfies , it can be found that the maximum value of Eq. (15) is

 Fmaxθ=4sin2(Bt2), (16)

which can be saturated when and , namely, the optimal state to access maximum quantum Fisher information here is a pure state perpendicular to , as shown in Fig. 1. In this figure, the yellow sphere represents the Bloch sphere and the blue arrow represents the vector . It can be found that all states on the joint ring of the green plane and surface of Bloch sphere can access the maximum quantum Fisher information, i.e., all states on this ring are optimal states. One simple example is , and another one is the superposition state of two eigenstates of  Giovannetti06 (); Pang ().

Alternatively, could be the parameter that under estimation. In the spin-half case, with respect to , , then the quantum Fisher information can be expressed by

 FB=t2|rin|2[1−(n0⋅re)2]. (17)

The optimal states to access the maximum value are the pure states vertical to .

Exponential form initial state. For an exponential form initial state , the parametrized state reads

 ρα=Uρ0U†=exp(UG0U†). (18)

Recently, Jiang Jiang14 () studied the quantum Fisher information for exponential states and gave a general form of SLD operator. In his theory, the SLD operator can be expanded as

 L=∞∑n=0gnG×n(∂αG), (19)

where the coefficient

 gn=4(2n+2−1)Bn+2(n+2)! (20)

for even and vanishes for odd . Here is the th Bernoulli number and in our case, . Through some straightforward calculation, the derivative of on reads

 ∂αG=−iU(G×0H)U†. (21)

Based on this equation, the th order term in Eq. (19) is

 G×n(∂αG)=−iU[(G×0)n+1H]U†, (22)

where is given by Eq. (9). Generally, it is known that the quantum Fisher information reads

 F=Tr(Uρ0U†L2)=Tr(ρ0L2eff), (23)

where the effective SLD operator . The effective SLD operator for pure states is already shown in Eq. (6). Substituting Eq. (22) into Eq. (19), the effective SLD operator can be expanded as

 Leff=−i∞∑n=0gn(G×0)n+1H. (24)

In most mixed states cases, to obtain quantum Fisher information, the diagonalization of initial state is inevitable, which is the reason why the usual form of quantum Fisher information is expressed in the eigenbasis of density matrix. Thus, it is worth to study the expression of effective SLD operator and quantum Fisher information in the eigenbasis of . We denote the th eigenvalue and eigenstate of as and , and in the eigenbasis of , the element of satisfies the recursion relation

 [G×n0H]ij=(ai−aj)[(G×0)n−1H]ij, (25)

where . Utilizing this recursive equation, a general formula of th order term can be obtained,

 [G×n0H]ij=(ai−aj)nHij. (26)

Substituting above equation into the expression of and being aware of the equality

 ∞∑n=0gn(ai−aj)n+1=2tanh(ai−aj2), (27)

the element of effective SLD operator in Eq. (24) can be written as

 [Leff]ij=−i2tanh(ai−aj2)Hij. (28)

Based on the equation , the quantum Fisher information in the eigenbasis of can finally be expressed by

 F=∑i>j4(eai+eaj)tanh2(ai−aj2)|Hij|2. (29)

This is one of the main results in this paper. In some real problems, the eigenspace of could be find easily. For instance, the eigenspace of a bosonic thermal state is the Fock space. Thus, as long as the formula of in Fock space is established, the quantum Fisher information can be obtained from Eq. (29).

Now we exhibit Eq. (29) with a spin-half thermal state. The initial state is taken as

 ρ0 = 1Zexp(−βσz), (30) = exp(−βσz−lnZ),

where with the Boltzmann constant and the temperature. In Planck unit, . The partition function reads . In this case, . Denoting the eigenstates of as and , i.e., , the eigenvalues of read and . The parametrization process is still taken as with the parameter under estimation, indicating that then the squared norm of the off-diagonal element of in the eigenbasis of reads

 |H01|2=sin2(Bt2)[1−cos2θcos2(Bt2)]. (31)

Immediately, the quantum Fisher information can be obtained from Eq. (29) as

 FT=4tanh2βsin2(Bt2)[1−cos2θcos2(Bt2)]. (32)

The maximum value of above expression is obtained at for and

 FmaxT=4tanh2(β). (33)

From this equation, one can see that the value of maximum quantum Fisher information is only affected by the temperature. With the increase of temperature, the maximum value reduces. In the other hand, quantum Fisher information in Eq. (32) is related to and . Fig. 2 shows the quantum Fisher information as a function of and . The values of and are both within in the plot. The temperature is set as here. From this figure, it can be found that the maximum quantum Fisher information is robust for since it is always obtained at for any value of . Furthermore, this optimal condition of is independent of temperature. With respect to , there is a large regime near in which the quantum Fisher information’s value can surpass , indicating that the quantum Fisher information can be still very robust and near its maximum value even when is hard to set exactly at .

Multiparameter processes. For a multiparameter system, the element of quantum Fisher information matrix in Ref. PhysicaAliu () can also be written with operator,

 Fαβ = M∑i=14picovi(Hα,Hβ) (34) −∑i≠j8pipjpi+pjRe(⟨ψi|Hα|ψj⟩⟨ψj|Hβ|ψi⟩),

where is dependent on a series of parameters , and so on, and

 Hm=i(∂mU†)U, (35)

with the index . The covariance matrix on the th eigenstate of initial state is defined as

 covi(Hα,Hβ):=12⟨ψi|{Hα,Hβ}|ψi⟩−⟨ψi|Hα|ψi⟩⟨ψi|Hβ|ψi⟩,

with the anti-commutation. For a single qubit system, Eq. (34) reduces to

 Fqubit,αβ=4(2Trρ2−1)cov1(2)(Hα,Hβ). (36)

Similarly with the single-parameter scenario, the subscript in Eq. (36) can be chosen as 1 or 2 since the covariance for two Hermitian operators are the same on two orthonormal states in 2-dimensional Hilbert space. From this equation, the element of quantum Fisher information matrix for pure states can be immediately obtained as

 Fpure,αβ=4covin(Hα,Hβ), (37)

namely, for pure states, the element of quantum Fisher information matrix is actually the covariance between two operators on the initial state. When the total Hamiltonian can be written as and for any , , above equation can reduce to the covariance between and  Liu_H (). For the diagonal elements, they are exactly the quantum Fisher information for the corresponding parameters.

For multiparamter estimations, the Cramér-Rao bound cannot always be achieved. In the scenario of pure states, the condition of this bound to be tight is ,  Matsumoto (); Fujiwara (). Here is dependent on the parameter under estimation. In the unitary parametrization, and this condition can be rewritten into , . Here is the effective SLD operator for parameter . Utilizing Eq. (6), this condition can be expressed in the form of operator,

 ⟨ψ0|[Hα,Hβ]|ψ0⟩=0,∀α,β. (38)

In other word, needs to be a real number for any and . When commutes with for any and , above condition can always be satisfied for any initial state.

Generally, for the unitary parametrization process, the element of quantum Fisher information matrix can be expressed by . From the definition equation of SLD, one can see that satisfies the equation . The quantum Fisher information matrix has more than one definitions. One alternative candidate is using the so-called Right Logarithmic Derivative (RLD) Holevo (); Belavkin (); Kim (), which is defined as , with the RLD. The element of RLD quantum Fisher information matrix can be written as

 Jαβ=Tr(ρRαR†β)=Tr(ρ0RαeffRβ†eff), (39)

where the effective RLD reads . For a unitary parametrization process, assuming the initial state has nonzero determinant, can be expressed by and the initial state , i.e.,

 Rαeff=i(ρ−10Hαρ0−Hα). (40)

With this equation, the element of RLD quantum Fisher information matrix can be expressed by

 (41)

When the parametrization process is displacement, this equation can reduces to the corresponding form in Ref. Kim (). For pure states, the element reads . Recently, Genoni et al. Kim () proposed a most informative Cramér-Rao bound for the total variance of all parameters under estimation. From the relation between and , one can see that is always larger than , namely, the SLD Cramér-Rao bound is always more informative than the RLD counterpart in this scenario.

We still consider the spin-half system with the Hamiltonian . Take both and as the parameters under estimations. First, based on aforementioned calculation, the operator for and read

 HB = −t2n0⋅σ, (42) Hθ = ∣∣∣sin(Bt2)∣∣∣n1⋅σ. (43)

Based on the property of Pauli matrices , the anti-commutation in the covariance reads

 {HB,Hθ}=−t∣∣∣sin(Bt2)∣∣∣n0⋅n1. (44)

For a pure initial state, the off-diagonal element of the quantum Fisher information matrix is expressed by

 FBθ=2t∣∣∣sin(Bt2)∣∣∣(n0⋅rin)(n1⋅rin), (45)

where is the Bloch vector of the initial pure state and the equality has been used. When the initial pure state is vertical to or , this off-diagonal element vanishes. Compared with the optimal condition for maximum quantum Fisher information for and individually, the Bloch vector can optimize both the diagonal elements of quantum Fisher information matrix and vanish the off-diagonal elements. However, all above is only necessary conditions for the achievement of Cramér-Rao bound. To find out if the bound can be really achieved, the condition (38) needs to be checked. In this case,

 [HB,Hθ]=−it∣∣∣sin(Bt2)∣∣∣n2⋅σ. (46)

With this equation, condition (38) reduces to , i.e., to make the Cramér-Rao bound achievable, the Bloch vector of the initial state needs to in the plane of and . Unfortunately, is not in this plane. Thus, and cannot be optimally joint measured simultaneously.

In the plane constructed by and , any Bloch vector of pure state can be written as , then we have , , and . From these expressions, one can see that the determinant of quantum Fisher information matrix is zero, i.e., . This fact indicates that, utilizing any pure state in this plane, the variances of and cannot be estimated simultaneously through the Cramér-Rao theory.

Discussion
We have discussed the quantum Fisher information with unitary parametrization utilizing an alternative representation. The total information of the parametrization process is involved in a operator in this representation. This operator is totally determined by the parameter and parametrization transformation . As long as the parameter and transformation are taken, is a settled operator and independent of the initial state. More interestingly, can be expressed in an expanded form. For the Hamiltonians owning recursive commutations with their partial derivative on the parameter under estimation, this expanded form shows a huge advantage. Utilizing this representation, we give a general analytical expression of quantum Fisher information for an exponential form initial state. Moreover, we have also studied the representation in multiparameter processes. The condition of Cramér-Rao bound to be achievable for pure states are also presented in the form of operator. In addition, we give the representation of Right Logarithmic Derivative and the corresponding quantum Fisher information matrix.

As a demonstration, we apply this representation in a collective spin system and show the expression of . Furthermore, we provide an analytical expression of quantum Fisher information in a spin-half system. If we consider this system as a multiparameter system, the corresponding quantum Fisher information matrix can also be straightforwardly obtained by this representation. From these expressions, one can find the optimal states to access the maximum quantum Fisher information. For the parameter , the optimal state is a pure state vertical to , and for the parameter , the optimal one is also a pure state, but vertical to . By analyzing the off-diagonal element of quantum Fisher information matrix, the states to optimize the diagonal elements and make the off-diagonal elements vanish are found. However, these states fail to satisfy the condition of achievement. Thus, and cannot be optimally jointed measured.

Methods
Collective spin system in a magnetic field.
For the Hamiltonian (12), its derivative on parameter is with the vector . Based on Eq. (9), can be written as

 H=[exp(itH×θ)−1]Jy. (47)

It is worth to notice that , then is

 H=[exp(iBtJ×n0)−1]Jy. (48)

Being aware of the commutation relations

 [Jn0,Jy] = iJn′0, (49) [Jn0,Jn′0] = −iJy, (50)

one can straightforwardly obtain the th order term as below

 J×nn0Jy={Jy,forevenn;iJn′0.foroddn. (51)

With this equation, can be expressed by

 H=[cos(Bt)−1]Jy−sin(Bt)Jn′0, (52)

equivalently, it can be written in a inner product form: , where the elements of read , and . After the normalization process, is rewritten into the form of Eq. (13).

For a spin-half system, the quantum Fisher information can be expressed by

 F=4sin2(Bt2)|rin|2[1−(n1⋅⟨σ⟩1(2))2], (53)

where is the Bloch vector of and can be obtained through the equation

 ρ0=12\openone+12∑i=x,y,zrin,iσi, (54)

with the identity matrix. is the vector of expected values on the th () eigenstate of . It can also be treated as the Bloch vector of the eigenstates. In previous sections, we denote .

## References

• (1) Napolitano, M. et al. Interaction-based quantum metrology showing scaling beyond the Heisenberg limit. Nature 471, 486–489 (2011); DOI: 10.1038/nature09778.
• (2) Riedel, M.F. et al. Atom-chip-based generation of entanglement for quantum metrology. Nature 464, 1170-1173 (2010); DOI: 10.1038/nature08988.
• (3) Strobel, H. et al. Fisher information and entanglement of non-Gaussian spin states. Science 345, 424-427 (2014); DOI: 10.1126/science.1250147.
• (4) Nagata, T. et al. Beating the Standard Quantum Limit with Four-Entangled Photons, Science 316, 726-729 (2007); DOI: 10.1126/science.1138007.
• (5) Pezze, L. & Smerzi, A. Ultrasensitive Two-Mode Interferometry with Single-Mode Number Squeezing. Phys. Rev. Lett. 110, 163604 (2013); DOI: http://dx.doi.org/10.1103/PhysRevLett.110.163604.
• (6) Gerry, C.C. & Mimih, J. The parity operator in quantum optical metrology. Contemp. Phys. 51, 497 (2010); DOI: 10.1080/00107514.2010.509995.
• (7) Anisimov, P.M. et al. Quantum Metrology with Two-Mode Squeezed Vacuum: Parity Detection Beats the Heisenberg Limit. Phys. Rev. Lett. 104, 103602 (2010); DOI: http://dx.doi.org/10.1103/PhysRevLett.104.103602.
• (8) Chin, A.W., Huelga, S.F. & Plenio, M.B., Quantum Metrology in Non-Markovian Environments. Phys. Rev. Lett. 109, 233601 (2012); DOI: http://dx.doi.org/10.1103/PhysRevLett.109.233601.
• (9) Genoni, M.G., Olivares, S. & Paris, M.G.A., Optical Phase Estimation in the Presence of Phase Diffusion. Phys. Rev. Lett. 106, 153603 (2011); DOI: http://dx.doi.org/10.1103/PhysRevLett.106.153603.
• (10) Feng, X.M., Jin, G.R. & Yang W., Quantum interferometry with binary-outcome measurements in the presence of phase diffusion. Phys. Rev. A 90, 013807 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.90.013807.
• (11) Lu, X.-M. et al. Broadcasting quantum Fisher information. Phys. Rev. A 87, 050302(R) (2013); DOI: http://dx.doi.org/10.1103/PhysRevA.87.050302.
• (12) Liu, J., Jing, X. & Wang, X. Phase-matching condition for enhancement of phase sensitivity in quantum metrology. Phys. Rev. A 88, 042316 (2013); DOI: http://dx.doi.org/10.1103/PhysRevA.88.042316.
• (13) Yao, Y., Ge, L., Xiao, X., Wang, X. & Sun, C.P. Multiple phase estimation in quantum cloning machines. Phys. Rev. A 90, 022327 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.90.022327.
• (14) Yao, Y. et al. Quantum Fisher information in noninertial frames. Phys. Rev. A 89, 042336 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.89.042336.
• (15) Yao, Y. et al. Multiple phase estimation for arbitrary pure states under white noise. Phys. Rev. A 90, 062113 (2014). DOI: http://dx.doi.org/10.1103/PhysRevA.90.062113.
• (16) Xiao, X., Yao, Y., Zhou, L.-M. & Wang, X. Distribution of quantum Fisher information in asymmetric cloning machines. Scientific Reports 4, 7361 (2014); DOI: 10.1038/srep07361.
• (17) Yue, J.-D., Zhang, Y.-R. & Fan, H. Quantum-enhanced metrology for multiple phase estimation with noise. Scientific Reports 4, 5933 (2014); DOI: 10.1038/srep05933.
• (18) Lang, M.D. & Caves, C.M. Optimal quantum-enhanced interferometry. Phys. Rev. A 90, 025802 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.90.025802.
• (19) Escher, B.M., de Matos Filho, R.L. & Davidovich, L. General framework for estimating the ultimate precision limit in noisy quantum-enhanced metrology. Nat. Phys. 7, 406-411 (2011); DOI: 10.1038/nphys1958.
• (20) Vidrighin, M.D. et al. Joint estimation of phase and phase diffusion for quantum metrology. Nat. Commun. 5, 3532 (2014); DOI: 10.1038/ncomms4532.
• (21) Braun, D. & Martin, J. Heisenberg-limited sensitivity with decoherence-enhanced measurements. Nat. Commun. 2, 223 (2014); DOI: 10.1038/ncomms1220.
• (22) Giovannetti, V., Llody, S. & Maccone, L. Quantum Metrology. Phys. Rev. Lett. 96, 010401 (2006); DOI: http://dx.doi.org/10.1103/PhysRevLett.96.010401.
• (23) Boixo, S., Flammia, S.T., Caves, C.M. & Geremia, J.M. Generalized Limits for Single-Parameter Quantum Estimation. Phys. Rev. Lett. 98, 090401 (2007); DOI: 10.1103/PhysRevLett.98.090401.
• (24) Taddei, M. M., Escher, B. M., Davidovich, L. & de Matos Filho, R. L. Quantum Speed Limit for Physical Processes. Phys. Rev. Lett. 110, 050402 (2013); DOI: 10.1103/PhysRevLett.110.050402.
• (25) Helstrom, C. W. Quantum Detection and Estimation Theory (Academic Press, New York, 1976).
• (26) Holevo, A. S. Probabilistic and Statistical Aspects of Quantum Theory (North-Holland, Amsterdam, 1982).
• (27) Pang, S. & Brun, T. Quantum metrology for a general Hamiltonian parameter. Phys. Rev. A 90, 022117 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.90.022117.
• (28) Jiang Z. Quantum Fisher information for states in exponential form. Phys. Rev. A 89, 032128 (2014); DOI: http://dx.doi.org/10.1103/PhysRevA.89.032128.
• (29) Liu, J., Jing, X., Zhong, W. & Wang, X. Quantum Fisher information for density matrices with arbitrary ranks. Commun. Theor. Phys. 61, 45 (2014); DOI: 10.1088/0253-6102/61/1/08.
• (30) Liu, J., Xiong, H.-N., Song, F. & Wang, X. Fidelity susceptibility and quantum Fisher information for density operators with arbitrary ranks. Phyica A 410, 167 (2014); DOI: 10.1016/j.physa.2014.05.028.
• (31) Liu, J., Lu, X.-M., Sun, Z. & Wang, X. Quantum multiparameter metrology with generalized entangled coherent state E-print arXiv: 1409.6167.
• (32) Matsumoto, K. A new approach to the Cramer-Rao-type bound of the pure-state model. J. Phys. A 35, 3111 (2002); DOI: 10.1088/0305-4470/35/13/307.
• (33) Fujiwara, A. Estimation of SU(2) operation and dense coding: An information geometric approach. Phys. Rev. A 65, 012316 (2001); DOI: http://dx.doi.org/10.1103/PhysRevA.65.012316.
• (34) Belavkin, V. P. Generalized uncertainty relations and efficient measurements in quantum systems Theor. Math. Phys. 26, 316 (1976); DOI: http://link.springer.com/article/10.1007/BF01032091.
• (35) Genoni, M. G. et al. Optimal estimation of joint parameters in phase space. Phys. Rev. A 87, 012107 (2013); DOI: http://dx.doi.org/10.1103/PhysRevA.87.012107.

Acknowledgments
The authors thank Dr. X.-M. Lu for helpful discussion. This work was supported by the NFRPC through Grant No. 2012CB921602 and the NSFC through Grants No. 11475146.

Author contributions
X.W. and J.L. contributed the idea. J.L. performed the calculations and prepared the figures. X.J. checked the calculations. J.L. wrote the main manuscript and X.W. made an improvement. All authors contributed to discussion and reviewed the manuscript.

Competing financial interests: The authors declare no competing financial interests.

You are adding the first comment!
How to quickly get a good reply:
• Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
• Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
• Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters