NDS Chernoff coefficients including support endpoints
OpenQAlgorithms.Papers.Independent.Intakethird100.P54.nds_chernoffLet be the finite number of signal modes and the finite number of retained idler modes, on the full optical Fock spaces. The two channels have equal priors, independent vacuum environments, reflectances , field magnitudes and , and distinct complex reflection coefficients . Every signal mode sees the same channel; idlers are retained losslessly and environment outputs are inaccessible. Write for squared Uhlmann fidelity, for minimum average error over all joint binary receivers, , and , with zero powers interpreted as support projections.
For a normalized pure NDS probe, let and denote its actual loss-sector probabilities and overlaps. For every common exponent the following series converges:
Sectors with either probability zero contribute zero. There exists attaining , and . In particular, endpoint powers are support projections, so the formula does not replace or by one.
The expression yields computable Chernoff bounds from the same NDS sector data used for exact discrimination.
import Definitions.Def_QAlgorithms_Papers_Independent_Intakethird100_P54
noncomputable section
open scoped BigOperators ComplexConjugate
open MeasureTheory Filter
namespace QAlgorithms.Papers.Independent.Intakethird100.P54
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Claimed50.P47 (Occupation)
open QAlgorithms.Papers.Independent.Dependency50.P39 (coherentCoefficient)
attribute [local instance] Classical.propDecidable
attribute [local instance] Classical.decEq
/-- NDS Chernoff-type formula with support endpoints and omitted empty sectors,
PDF p.6, (45), together with (7)–(9). The optimizer is concluded, not assumed. -/
theorem nds_chernoff {M J : ℕ} (hM : 0 < M) (c : Channels) (hc : ValidSplitters c)
(psi : Hilbert (Occupation (Fin J) × Occupation (Fin M))) (hpsi : ‖psi‖ = 1)
(hnds : IsNDS psi) :
(∀ s ∈ Set.Icc (0 : ℝ) 1, Summable (ndsChernoffTerm c psi s) ∧
outputChernoffAt c psi s = ∑' k, ndsChernoffTerm c psi s k ∧
outputError c psi ≤ outputChernoffAt c psi s / 2) ∧
(∃ s ∈ Set.Icc (0 : ℝ) 1, outputChernoff c psi = outputChernoffAt c psi s ∧
∀ t ∈ Set.Icc (0 : ℝ) 1, outputChernoffAt c psi s ≤ outputChernoffAt c psi t) ∧
outputError c psi ≤ outputChernoff c psi / 2 := by sorry
end QAlgorithms.Papers.Independent.Intakethird100.P54
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let with , allowing , and let have norm and satisfy whenever . For , let , , , and , with and . Define for componentwise, and zero otherwise; put , let equal this function when square-summable and zero otherwise, and set , , and . A resolution of a kernel means summable nonnegative weights of sum , vectors unit at positive weights and orthogonal at distinct positive weights, and , with zero-weight vectors unrestricted. Put when , and , using real powers, otherwise. Define
where the resolutions are of . Define as the infimum, over these resolutions and bounded complex-linear effects , of , where , , and for every vector. The real suprema and infima are total, taking value zero on empty defining sets. For each , put
The theorem asserts that for every , the sequence is summable, , and ; that some satisfies and for every ; and that . Both endpoints are included, and a zero resolution weight contributes zero even when its exponent is zero.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.