Uncertainty Relations in Unitary Implementation

• DESIRED Relations are uncertainty trade-offs showing that precise implementation of energy-changing unitaries requires significant apparatus energy fluctuation and quantum coherence.
• The framework analyzes closed joint dynamics under energy conservation using Bures-type and entanglement-fidelity metrics to quantify implementation error.
• Implications include scaling challenges for macroscopic systems and limitations on classical control, necessitating unbounded coherent energy variance for near-perfect operation.

Searching arXiv for the target paper and closely related context papers. “Uncertainty relations in implementation of unitary operations” studies the physical realization of a target unitary transformation on a quantum system when the transformation is implemented through interaction with an external apparatus rather than postulated as an abstract gate. In this formulation, the system $S$ and apparatus $E$ undergo closed joint unitary dynamics generated by a time-independent total Hamiltonian, while the reduced dynamics on $S$ approximates a desired unitary $U_S$. The central result is a set of uncertainty-type trade-off relations: if $U_S$ does not commute with the system Hamiltonian $H_S$, then high-accuracy implementation requires large initial energy fluctuation in the apparatus, and, more strongly, a large quantum-coherent component of that fluctuation (Tajima et al., 2017).

1. Physical setting and operational definition

The framework considers a system $S$ with Hilbert space $\mathcal H_S$ and Hamiltonian $H_S$, together with an external apparatus $E$ with Hilbert space $E$0 and Hamiltonian $E$1. The apparatus starts in a state $E$2, which may be pure or mixed. The intended operation on $E$3 is a fixed unitary $E$4, while the actual implementation is induced by closed evolution of $E$5 under

$E$6

so that at time $E$7 the joint dynamics is

$E$8

and the reduced map on $E$9 is

$S$0

This reduced map is CPTP, whereas the composite $S$1 dynamics is closed and unitary (Tajima et al., 2017).

The main results assume energy conservation in the sense

$S$2

A generalized setting allows weak violation of this condition, measured by

$S$3

which weakens but does not eliminate the resulting bounds as long as $S$4 remains smaller than the energy change implied by the target (Tajima et al., 2017).

Operationally, an implementation is specified by

$S$5

A “good” implementation is one for which $S$6 is close to the target unitary channel generated by $S$7. This setup makes the resource question sharply physical: a unitary gate is not treated as an abstract primitive, but as a transformation that must emerge from a controller or apparatus with its own Hamiltonian and quantum state (Tajima et al., 2017).

2. Accuracy measures and energy-fluctuation quantifiers

The implementation error is defined through a worst-case state approximation metric based on a Bures-type distance. The paper defines

$S$8

with

$S$9

and then

$U_S$0

Notably, this definition inserts an additional square root, so $U_S$1 is the maximal Bures-type distance appearing in the paper’s convention (Tajima et al., 2017).

A second, entanglement-sensitive accuracy measure is also introduced. It is based on entanglement fidelity and the corresponding entanglement-Bures length:

$U_S$2

where

$U_S$3

This variant yields stronger constants in the trade-off relations, but uses a different metric (Tajima et al., 2017).

On the resource side, the first relevant quantity is the total energy standard deviation in the apparatus initial state,

$U_S$4

The second is the quantum-coherent component of that fluctuation, defined by a convex-roof construction over pure-state decompositions of $U_S$5:

$U_S$6

This quantity vanishes for mixtures of energy eigenstates and equals $U_S$7 for pure apparatus states. The paper identifies $U_S$8 as the fluctuation arising from quantum superposition across energy eigenstates and notes that it is equivalent, up to a known constant factor, to the quantum Fisher information for phase shifts generated by $U_S$9 (Tajima et al., 2017).

A further key quantity is the target-dependent incompatibility measure

$U_S$0

The paper gives the operational expression

$U_S$1

so it quantifies the maximum possible change in system energy under the desired unitary. This makes the subsequent bounds explicitly about energy-changing gates, rather than about unitary control in general (Tajima et al., 2017).

3. Main uncertainty relations

Under exact energy conservation, the primary bound relates total apparatus energy fluctuation to implementation accuracy. For sufficiently accurate implementations satisfying

$U_S$2

the paper proves

$U_S$3

If $U_S$4, then arbitrarily accurate implementation forces $U_S$5 to grow, and perfect implementation with vanishing apparatus energy fluctuation is impossible unless the target commutes with the system Hamiltonian (Tajima et al., 2017).

The more distinctive result concerns the coherent part of the apparatus fluctuation. For implementations satisfying

$U_S$6

one has

$U_S$7

This excludes high-fidelity implementation by an apparatus that is merely a classical mixture over energy eigenstates. In the paper’s formulation, coherent superposition across many energy levels with broad support is necessary for accurate realization of an energy-changing unitary (Tajima et al., 2017).

When energy conservation is only approximate, the same structure persists with an effective incompatibility reduced by $U_S$8. If

$U_S$9

then

$H_S$0

Thus, so long as the degree of energy non-conservation is smaller than the energy change implied by the target, the essential resource-pressure remains (Tajima et al., 2017).

Using the entanglement-fidelity metric yields a stronger constant. If

$H_S$1

then

$H_S$2

The paper does not claim the constants $H_S$3, $H_S$4, and $H_S$5 are optimal, and explicitly notes that equalities are generally unattainable except in the trivial commuting case $H_S$6 (Tajima et al., 2017).

4. Derivation strategy and conceptual structure

The derivation combines two inequalities. The first is a variance–distance relation for a Hermitian observable $H_S$7 and states $H_S$8. Writing

$H_S$9

and

$S$0

the paper uses

$S$1

This bounds expectation-value differences in terms of variance and Bures distance (Tajima et al., 2017).

The second ingredient is a no-information back-action statement for the apparatus. For orthogonal pure inputs $S$2 and $S$3, let the corresponding apparatus outputs be $S$4 and $S$5. If $S$6, then

$S$7

Hence, accurate implementation implies that the apparatus marginal is almost independent of which orthogonal input state was used, so the apparatus cannot significantly record which-input information (Tajima et al., 2017).

The proof then chooses system inputs that maximize and minimize the energy change under $S$8, namely eigenvectors of $S$9. Combining the two inequalities with conservation of $\mathcal H_S$0 links three quantities: the apparatus expectation change in $\mathcal H_S$1 across the two runs, the post-interaction apparatus energy variances, and the commutator norm $\mathcal H_S$2, up to corrections of order $\mathcal H_S$3. Rearrangement yields the uncertainty-type inequalities (Tajima et al., 2017).

For the $\mathcal H_S$4 bound, the same argument is repeated on each pure component of an ensemble decomposition of $\mathcal H_S$5, followed by minimization over all decompositions. This convex-roof step isolates the genuinely quantum part of the resource, excluding classical mixing as a substitute for coherent energy spread (Tajima et al., 2017).

The paper explicitly situates these results as an implementation analogue of Wigner–Araki–Yanase limitations and as echoing Ozawa-type measurement-disturbance trade-offs, but with the object of study shifted from measurement to unitary control (Tajima et al., 2017).

5. Scaling, examples, and physical interpretation

The dependence on $\mathcal H_S$6 makes the bounds sensitive to the energetic noncommutativity of the target operation. If $\mathcal H_S$7 significantly reshuffles energy under an extensive Hamiltonian $\mathcal H_S$8, then $\mathcal H_S$9 can scale with system size $H_S$0. For fixed target accuracy $H_S$1, the paper therefore identifies the necessary scaling

$H_S$2

so the required apparatus energy fluctuation is at least extensive for macroscopic energy-changing unitaries (Tajima et al., 2017).

A direct asymptotic consequence is that, when $H_S$3 and $H_S$4, both $H_S$5 and $H_S$6 must diverge. This suggests that near-perfect implementation requires asymptotically unbounded apparatus energy fluctuation and coherent bandwidth (Tajima et al., 2017).

The paper illustrates the bounds with the Jaynes–Cummings model, using

$H_S$7

and a coherent apparatus state $H_S$8 with

$H_S$9

In the limit $E$0 with $E$1 fixed, the system evolves under

$E$2

For the choice $E$3, one gets $E$4, so the main bound implies

$E$5

The paper gives the numerical example $E$6 and $E$7, which yields $E$8 and therefore $E$9 (Tajima et al., 2017).

By contrast, if $E$00, for example when $E$01 is a function of $E$02, the bounds become vacuous. In that commuting case, zero apparatus energy fluctuation is compatible with perfect implementation (Tajima et al., 2017).

6. Scope, limitations, and significance

The results are model-independent within the stated assumptions. They hold for arbitrary apparatus Hilbert space and Hamiltonian, arbitrary interaction Hamiltonian $E$03, and arbitrary target $E$04, provided the closed dynamics is unitary and the total evolution satisfies exact or approximate conservation of $E$05 (Tajima et al., 2017).

Several limitations are explicit. The constants are not optimized; the small-error conditions are sufficient rather than necessary; the bounds do not contain the gate time $E$06 except indirectly through $E$07 and $E$08; and the analysis addresses single-shot, one-step implementations, although the paper states that multi-step and feedback schemes still inherit the same resource-pressure through the net target and energy conservation (Tajima et al., 2017).

The practical interpretation is direct. For energy-changing gates such as transverse rotations on a Zeeman qubit, low error demands an apparatus with large energy fluctuation. Moreover, the fluctuation must be quantum-coherent rather than merely classical. The paper states that classical mixtures lead to dissipation rather than coherent control, whereas coherent superpositions with broad energy support—such as highly populated coherent fields or other energy-spread states—supply the required resource (Tajima et al., 2017).

A common misconception is that increasing control precision merely requires stronger classical driving amplitude. The trade-off with $E$09 shows that, under energy conservation, accuracy is tied not just to apparatus energy variance in the aggregate, but to its coherent component. Another misconception is that the result is restricted to a specific model such as Jaynes–Cummings; in fact, the paper emphasizes universality across apparatus dimension, interaction form, and target operation (Tajima et al., 2017).

In that sense, the work converts an intuitive control-theoretic idea into a quantitative impossibility statement: when a desired unitary changes system energy, precise coherent implementation is not free. It consumes apparatus-side quantum energy fluctuation as a fundamental resource, and the required amount diverges in the limit of perfect accuracy (Tajima et al., 2017).