You are handed one copy of a state that is either ρ0 or ρ1. How well can you guess which? The answer, the Helstrom bound, takes one line, and it anchors a whole theory: many hypotheses, error-free (unambiguous) discrimination, many copies, and the quantum Chernoff and Stein exponents.
What you’ll learn
Derive the Helstrom–Holevo bound and construct the optimal measurement.
Formulate multi-hypothesis discrimination as a semidefinite program, with the Holevo–Yuen–Kennedy–Lax optimality conditions and the pretty good measurement.
Prove the Ivanovic–Dieks–Peres bound for unambiguous discrimination.
State the asymptotic error exponents: the quantum Chernoff bound and quantum Stein's lemma.
Two hypotheses: the Helstrom–Holevo bound
A source emits ρ0 with prior probability p0, or ρ1 with p1=1−p0. You measure a two-outcome POVM{E0,E1=I−E0} and guess according to the outcome.
Everything depends on the Hermitian Helstrom operatorΓ.
0≤E0≤ImaxTr(E0Γ)=TrΓ+
Take E0 to be the projector onto the positive eigenspace of Γ, exactly as in the proof that trace distance is the best bias.
TrΓ+=21(∥Γ∥1+TrΓ)=21(∥Γ∥1+p0−p1)
Because Γ=Γ+−Γ− and ∥Γ∥1=TrΓ++TrΓ−.
Psuccopt=21(1+∥p0ρ0−p1ρ1∥1)
Using p1+21(p0−p1)=21. For equal priors this is 21(1+D(ρ0,ρ1)), with D the trace distance.
Predict before you look
Equal priors, ∣ψ0⟩=∣0⟩ and ∣ψ1⟩=cos2θ∣0⟩+sin2θ∣1⟩ with θ=60∘. What is the optimal success probability?
Show the answer
0.75
s=cos30∘=23, so P=21(1+1−43)=21(1+21)=43. On the Bloch sphere the two states are 60∘ apart, and D=21∣r0−r1∣=sin30∘=21.
For qubits with equal priors, the optimal measurement is the observable n^⋅σ with n^∝r0−r1: guess ρ0 on outcome +1. Priors matter, too. If Γ≥0, then ∥Γ∥1=p0−p1 and the optimum Psucc=p0 is reached by ignoring the system and always guessing ρ0: a measurement helps only when it can overturn the prior.
Many hypotheses: a semidefinite program
For m states ρi with priors pi, the optimal success probability is a semidefinite program (SDP), and its dual certifies optimality:
Weak duality takes one line: ∑ipiTr(Eiρi)≤∑iTr(EiZ)=TrZ, because Z−piρi≥0 and Ei≥0. Both problems are strictly feasible, so strong duality holds and the two values agree.
Check your understanding
Five equiprobable states of a qubit, otherwise arbitrary. What is the best upper bound, valid for every such ensemble, on the success probability of any measurement?
1
1/2
2/5
1/5
Show the answer
C.2/5
Psucc=51∑iTr(Eiρi)≤51∑iTrEi=51TrI=52. The bound is attained, for example, by five pure states equally spaced on a great circle of the Bloch sphere, measured with their PGM.
Unambiguous discrimination
DerivationProof of the bound
⟨ψ0∣ψ1⟩=⟨ψ0∣(E0+E1+E?)∣ψ1⟩=⟨ψ0∣E?∣ψ1⟩
Let {E0,E1,E?} be error-free: ⟨ψ1∣E0∣ψ1⟩=0, which forces E0∣ψ1⟩=0 because E0≥0; likewise E1∣ψ0⟩=0. Therefore
s2≤q0q1≤(2q0+q1)2
Apply Cauchy–Schwarz to the positive semidefinite form (u,v)↦⟨u∣E?∣v⟩, with qi=⟨ψi∣E?∣ψi⟩ the probability of an inconclusive result on ∣ψi⟩, then AM–GM:
With equal priors Psucc=1−21(q0+q1)≤1−s. □ Achievability: the three-outcome POVM of the Excited State lesson, realized by the two-qubit circuit in Naimark Dilation.
With unequal priors p0≤p1, the optimum is 1−2p0p1s as long as s≤p0/p1. Beyond that threshold it is better to give up on the less likely state and use a projective measurement that only ever identifies ∣ψ1⟩, succeeding with probability p1(1−s2) (Jaeger & Shimony 1995). For more than two states, pure states can be unambiguously discriminated, each with nonzero success probability, iff they are linearly independent (Chefles 1998).
Figure of merit
Optimum for ∣0⟩ vs ∣+⟩, equal priors
Optimal measurement
Minimum error (Helstrom)
success cos28π≈0.854, error ≈0.146
projective
Unambiguous (Ivanovic–Dieks–Peres)
success 1−21≈0.293, never wrong, inconclusive ≈0.707
three-outcome POVM
Unambiguous, projective only
success 0.25
measure Z (or X)
Many copies: error exponents
With n copies, ρi→ρi⊗n, and the optimal error falls exponentially in n. Two regimes are fully understood.
For pure states Tr(ρ0αρ11−α)=∣⟨ψ0∣ψ1⟩∣2 for every 0<α<1, so ξQCB=−log∣⟨ψ0∣ψ1⟩∣2. This matches the exact n-copy Helstrom error for equal priors, 21(1−1−s2n)≈41s2n, since n copies have overlap sn.
Attaining these exponents generally requires collective measurements across the copies. For two pure states, however, measuring one copy at a time with adaptive feed-forward (Bayesian updating of the prior) already attains the optimal n-copy error; for mixed states, such local strategies can fall short of the collective optimum.
Key takeaways
Helstrom–Holevo: Psuccopt=21(1+∥p0ρ0−p1ρ1∥1), achieved by projecting onto the positive part of p0ρ0−p1ρ1.
Many hypotheses: an SDP whose dual certificate Z≥piρi gives the Holevo–Yuen–Kennedy–Lax conditions; the pretty good measurement is within a square of optimal.
Unambiguous discrimination of two equiprobable pure states succeeds with probability at most 1−∣⟨ψ0∣ψ1⟩∣, which requires a genuine POVM.
With n copies, errors decay exponentially: with the Chernoff exponent −logminαTrρ0αρ11−α (symmetric) or the relative entropy (Stein, asymmetric).
Check yourself
In the interactive lesson these are scored with your confidence, and what you miss comes back for review.
Question 1
Equal priors, ρ0=∣0⟩⟨0∣ and ρ1=I/2. What is the optimal success probability?
1/2
3/4
cos28π≈0.854
1
Show the answer
B.3/4
D(∣0⟩⟨0∣,I/2)=21z^−0=21, so P=21(1+21)=43. Concretely: measure Z, guess ρ0 on 0 and ρ1 on 1.
Question 2
For two pure states with overlap s=∣⟨ψ0∣ψ1⟩∣<1 and equal priors, how does the minimum error probability for n copies decay?
As s2n up to a constant, i.e. with Chernoff exponent −log∣⟨ψ0∣ψ1⟩∣2
As sn
Polynomially, like 1/n
As (1−s)n
Show the answer
A. As s2n up to a constant, i.e. with Chernoff exponent −log∣⟨ψ0∣ψ1⟩∣2
The n-copy states have overlap sn, so Perr=21(1−1−s2n)≈41s2n, matching ξQCB=−logs2.
Question 3
Which statement about the pretty good measurement (PGM) is correct?
It is optimal for every ensemble
It always succeeds with probability at least 1/2
It is defined only for commuting states
It satisfies PsuccPGM≥(Psuccopt)2 for every ensemble
Show the answer
D. It satisfies PsuccPGM≥(Psuccopt)2 for every ensemble
The Barnum–Knill bound makes the PGM near-optimal whenever the optimum is close to 1. It is exactly optimal for geometrically uniform pure-state ensembles, but not in general.
Question 4
In quantum Stein's lemma, which quantity is the optimal exponent of the type-II error?
The trace distance D(ρ0,ρ1)
−logF(ρ0,ρ1)
The relative entropy D(ρ0∥ρ1)=Trρ0(logρ0−logρ1)
The Chernoff exponent ξQCB
Show the answer
C. The relative entropy D(ρ0∥ρ1)=Trρ0(logρ0−logρ1)
With the type-I error held below a constant, the type-II error decays as e−nD(ρ0∥ρ1). The Chernoff exponent governs the symmetric setting, where both errors are weighted by priors, and is never larger than the relative entropy.
Further reading
C. W. Helstrom, Quantum Detection and Estimation Theory (Academic Press, 1976).
A. S. Holevo, "Statistical decision theory for quantum systems", Journal of Multivariate Analysis 3, 337 (1973); H. P. Yuen, R. S. Kennedy & M. Lax, "Optimum testing of multiple hypotheses in quantum detection theory", IEEE Transactions on Information Theory 21, 125 (1975).
K. M. R. Audenaert, J. Calsamiglia, R. Muñoz-Tapia, E. Bagan, Ll. Masanes, A. Acín & F. Verstraete, "Discriminating states: the quantum Chernoff bound", Physical Review Letters 98, 160501 (2007).
H. Barnum & E. Knill, "Reversing quantum dynamics with near-optimal quantum and classical fidelity", Journal of Mathematical Physics 43, 2097 (2002).
S. M. Barnett & S. Croke, "Quantum state discrimination", Advances in Optics and Photonics 1, 238 (2009).