Unnumbered result (20): finite-dimensional Holevo capacity bound
OpenQAlgorithms.Papers.Independent.Gap48.P45.holevoCapacity_inputDimensionBoundAll physical Hilbert spaces are nonzero complex separable Hilbert spaces, with finite dimensions included. States are positive trace-class operators of trace one, channels are completely positive and trace preserving on the entire trace class, and logarithms are natural. Inner products are conjugate-linear in the first argument. Let both input and output spaces have positive finite dimensions. Define the single-letter Holevo capacity by , over discrete input ensembles. Then
This result bounds the single-letter classical information by the input dimension.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P45
noncomputable section
open Filter MeasureTheory
open scoped BigOperators ENNReal Topology
attribute [local instance] Classical.propDecidable Classical.decEq
attribute [local instance] IsStarNormal.instContinuousFunctionalCalculus
namespace QAlgorithms.Papers.Independent.Gap48.P45
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Dependency50.P33
(State IsTraceClass IsTraceClassChannel traceSize traceClosure rankOne productCoeff tensorApply)
open QAlgorithms.Papers.Independent.Frontier18.P01 (Channel Complementary traceFirst traceSecond)
open QAlgorithms.Papers.Independent.Frontier18.P07
(GeneralizedEnsemble averageCoefficients extendedEntropyNats extendedRelativeEntropyNats ensembleHolevo pureStates finiteEnsembleCapacity)
open QAlgorithms.Papers.Independent.Frontier18.P18
(operatorCoeff boundedOperator conjugateCoeff ReversibleOn IsometricallyEquivalent
IsParsevalFrame IsKrausRepresentation)
open QAlgorithms.Papers.Independent.Claimed50.P16 (tensorRankLE)
open QAlgorithms.Papers.Independent.Next46.P31 (positiveSquareRoot)
open QAlgorithms.Papers.Independent.Next46.P42 (minimumOutputEntropy)
/-- Equation (20), PDF p.21: finite-dimensional single-letter capacity bound,
in nats. Both physical spaces are positive finite-dimensional. -/
theorem holevoCapacity_inputDimensionBound
{I J : Type} [Fintype I] [Fintype J] [Nonempty I] [Nonempty J]
(Phi : Channel I J) :
(finiteEnsembleCapacity Phi.val Set.univ) ≤ ENNReal.ofReal (Real.log (Fintype.card I : ℝ)) := by sorry
end QAlgorithms.Papers.Independent.Gap48.P45
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be finite, nonempty types and let be a channel between their coefficient matrices. Here states are matrices admitting an expansion in -vectors with , having real, nonnegative quadratic forms on every finite coordinate set, and having summable real diagonal of sum . A channel preserves matrices admitting such expansions, is additive and complex homogeneous on them, and preserves states under blockwise amplification by the identity on every finite auxiliary coordinate space. Let be the infimum of the norm-product sums of these expansions, with empty infimum ; equip states with the topology generated by -balls. For matrices , set , using finite coordinate compressions in the relevant index type, where
Here and . For a Borel probability measure on input states, write . Consider only those with finite topological support and with a state as barycenter, meaning that every coordinate function is integrable and . Then
The supremum and integrals take values in , and the right side is the extended nonnegative real obtained from ; since , it equals , including the bound when .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.