Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Eqs.(15)–(17): explicit symplectic diagonalization of the returned covariances

Open
QAlgorithms.Papers.Independent.Intakenext100.P85.return_williamson

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let NS,NB≥0N_S,N_B\ge0NS​,NB​≥0 and 0<κ<10<\kappa<10<κ<1. Write S=2NS+1S=2N_S+1S=2NS​+1, Cq=2NS(NS+1)C_q=2\sqrt{N_S(N_S+1)}Cq​=2NS​(NS​+1)​, B=2NB+1B=2N_B+1B=2NB​+1, A=2κNS+BA=2\kappa N_S+BA=2κNS​+B, and

Δ=(A+S)2−4κCq2,x±=A+S±Δ2Δ,T=(X+X−X−X+),X+=diag⁡(x+,x+), X−=diag⁡(x−,−x−).\Delta=\sqrt{(A+S)^2-4\kappa C_q^2},\quad x_\pm=\sqrt{\frac{A+S\pm\Delta}{2\Delta}},\quad T=\begin{pmatrix}X_+&X_-\\X_-&X_+\end{pmatrix},\quad X_+=\operatorname{diag}(x_+,x_+),\ X_-=\operatorname{diag}(x_-,-x_-).Δ=(A+S)2−4κCq2​​,x±​=2ΔA+S±Δ​​,T=(X+​X−​​X−​X+​​),X+​=diag(x+​,x+​), X−​=diag(x−​,−x−​).

Then Δ>0\Delta>0Δ>0, the identity diagonalizes the target-absent covariance with spectrum (B/4,S/4)(B/4,S/4)(B/4,S/4), and TTT is symplectic and diagonalizes the target-present covariance with

νk=(−1)k(S−A)+Δ8,k=1,2,ΛRI(1)=Tdiag⁡(ν1,ν1,ν2,ν2)TT.\nu_k=\frac{(-1)^k(S-A)+\Delta}{8},\quad k=1,2,\qquad \Lambda_{RI}^{(1)}=T\operatorname{diag}(\nu_1,\nu_1,\nu_2,\nu_2)T^T.νk​=8(−1)k(S−A)+Δ​,k=1,2,ΛRI(1)​=Tdiag(ν1​,ν1​,ν2​,ν2​)TT.

All symplectic eigenvalues are at least 1/41/41/4. Symplectic means TΩTT=ΩT\Omega T^T=\OmegaTΩTT=Ω for Ω=⨁j=12(01−10)\Omega=\bigoplus_{j=1}^2\left(\begin{smallmatrix}0&1\\-1&0\end{smallmatrix}\right)Ω=⨁j=12​(0−1​10​). This supplies the explicit physical diagonalizations used to evaluate Gaussian overlaps.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P85
Formal statement
noncomputable section

open MeasureTheory Filter

open scoped BigOperators ENNReal Topology

attribute [local instance] Classical.propDecidable Classical.decEq

namespace QAlgorithms.Papers.Independent.Intakenext100.P85

open QAlgorithms.Papers.Independent.Claimed50.P19 (Cov)

open QAlgorithms.Papers.Independent.Renew46.P41 (coherentProductCoefficient)

/-- Source p.3, Eqs.(15)–(17), with the definition (9)–(10).
Source spectrum indices 1,2 become Fin 2 values 0,1. Delta>0 is a
conclusion, so all square-root denominators are meaningful. -/
theorem return_williamson (k NS NB : ℝ) (hk : 0 < k) (hk1 : k < 1)
    (hNS : 0 ≤ NS) (hNB : 0 ≤ NB) :
    0 < returnDiscriminant k NS NB ∧
    (∀ present : Bool, ∀ r : Fin 2, 1 / 4 ≤ returnSpectrum present k NS NB r) ∧
    G.Symplectic (1 : Cov 2) ∧
    returnCov false k NS NB = G.repeatedDiagonal (returnSpectrum false k NS NB) ∧
    G.Symplectic (williamsonT k NS NB) ∧
    returnCov true k NS NB = williamsonT k NS NB *
      G.repeatedDiagonal (returnSpectrum true k NS NB) * (williamsonT k NS NB).transpose := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P85

end
Source
Quantum Illumination with Gaussian States, arXiv:0810.0534v1, PDF p. 3, Eqs.(15)–(17): explicit symplectic diagonalization of the returned covariances
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

For all real 0<k<10<k<10<k<1 and NS,NB≥0N_S,N_B\ge0NS​,NB​≥0, define

S=2NS+1,B=2NB+1,A=2kNS+B,Q=2NS(NS+1),D=(A+S)2−4kQ2,S=2N_S+1,\quad B=2N_B+1,\quad A=2kN_S+B,\quad Q=2\sqrt{N_S(N_S+1)},\quad D=\sqrt{(A+S)^2-4kQ^2},S=2NS​+1,B=2NB​+1,A=2kNS​+B,Q=2NS​(NS​+1)​,D=(A+S)2−4kQ2​,

and, in coordinate order (0,0),(0,1),(1,0),(1,1)(0,0),(0,1),(1,0),(1,1)(0,0),(0,1),(1,0),(1,1), define

V0=14diag⁡(B,B,S,S),V1=14(A0kQ00A0−kQkQ0S00−kQ0S).V_0=\frac14\operatorname{diag}(B,B,S,S),\qquad V_1=\frac14 \begin{pmatrix} A&0&\sqrt{k}Q&0\\ 0&A&0&-\sqrt{k}Q\\ \sqrt{k}Q&0&S&0\\ 0&-\sqrt{k}Q&0&S \end{pmatrix}.V0​=41​diag(B,B,S,S),V1​=41​​A0k​Q0​0A0−k​Q​k​Q0S0​0−k​Q0S​​.

Let

ν(0)=(B/4,S/4),ν(1)=((A−S+D)/8,(S−A+D)/8),\nu^{(0)}=(B/4,S/4),\qquad \nu^{(1)}=((A-S+D)/8,(S-A+D)/8),ν(0)=(B/4,S/4),ν(1)=((A−S+D)/8,(S−A+D)/8), x+=A+S+D2D,x−=A+S−D2D,T=(x+0x−00x+0−x−x−0x+00−x−0x+),Ω=diag⁡ ⁣((01−10),(01−10)).x_+=\sqrt{\frac{A+S+D}{2D}},\qquad x_-=\sqrt{\frac{A+S-D}{2D}},\qquad T=\begin{pmatrix} x_+&0&x_-&0\\ 0&x_+&0&-x_-\\ x_-&0&x_+&0\\ 0&-x_-&0&x_+ \end{pmatrix}, \qquad \Omega=\operatorname{diag}\!\left( \begin{pmatrix}0&1\\-1&0\end{pmatrix}, \begin{pmatrix}0&1\\-1&0\end{pmatrix}\right).x+​=2DA+S+D​​,x−​=2DA+S−D​​,T=​x+​0x−​0​0x+​0−x−​​x−​0x+​0​0−x−​0x+​​​,Ω=diag((0−1​10​),(0−1​10​)).

The theorem asserts D>0D>0D>0; νr(p)≥1/4\nu^{(p)}_r\ge1/4νr(p)​≥1/4 for both p=0,1p=0,1p=0,1 and both mode indices r=0,1r=0,1r=0,1; and

I4ΩI4T=Ω,V0=diag⁡(ν0(0),ν0(0),ν1(0),ν1(0)),I_4\Omega I_4^{\mathsf T}=\Omega,\qquad V_0=\operatorname{diag}(\nu^{(0)}_0,\nu^{(0)}_0,\nu^{(0)}_1,\nu^{(0)}_1),I4​ΩI4T​=Ω,V0​=diag(ν0(0)​,ν0(0)​,ν1(0)​,ν1(0)​), TΩTT=Ω,V1=T diag⁡(ν0(1),ν0(1),ν1(1),ν1(1)) TT.T\Omega T^{\mathsf T}=\Omega,\qquad V_1=T\,\operatorname{diag}(\nu^{(1)}_0,\nu^{(1)}_0,\nu^{(1)}_1,\nu^{(1)}_1)\,T^{\mathsf T}.TΩTT=Ω,V1​=Tdiag(ν0(1)​,ν0(1)​,ν1(1)​,ν1(1)​)TT.

Zero signal energy and zero background energy are included; the asserted strict positivity of DDD applies also at those endpoints.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me