Definition 7: Space and Time Faults #
Statement #
In the context of fault-tolerant quantum error correction:
A space-fault (or space error) is a Pauli error operator that occurs on some qubit at a specific time during the procedure.
A time-fault (or measurement error) is an error where the result of a measurement is reported incorrectly (i.e., +1 is reported as -1 or vice versa).
A spacetime fault is a collection of space-faults and time-faults occurring at various locations and times during the procedure.
Convention: State mis-initialization faults are treated as time-faults that are equivalent to a perfect initialization followed by a space-fault.
Main Results #
SpaceFault: A Pauli error on a specific qubit at a specific timeTimeFault: A measurement error (bit-flip of outcome) at a specific measurement locationSpacetimeFault: A collection of space-faults and time-faultsSpacetimeFault.weight: The total number of individual faultsSpacetimeFault.empty: The fault-free configurationSpacetimeFault.compose: Composition (symmetric difference) of faults
Corollaries #
- Weight properties: empty has weight 0, single faults have weight 1
- Composition is commutative and involutive
- Space-only and time-only fault embeddings
Space Faults #
A space-fault (space error) is a single-qubit Pauli error at a specific qubit and time. The error is described by which qubit is affected, when it occurs, and what Pauli error is applied (X, Y, or Z, encoded as a nonzero element of ZMod 2 x ZMod 2 for the x and z components).
- qubit : Q
The qubit where the error occurs.
- time : T
The time at which the error occurs.
- xComponent : ZMod 2
The X-component of the Pauli error (1 if X or Y, 0 otherwise).
- zComponent : ZMod 2
The Z-component of the Pauli error (1 if Z or Y, 0 otherwise).
The error is nontrivial: at least one component is nonzero.
Instances For
Time Faults #
A time-fault (measurement error) is a single measurement whose outcome is reported incorrectly. In ZMod 2 encoding (0 -> +1, 1 -> -1), this corresponds to flipping the measurement outcome bit. The fault is identified by which measurement is affected.
Convention: State mis-initialization faults are also modeled as time-faults. For example, initializing |0> but getting |1> is equivalent to a perfect initialization followed by an X error. The initialization measurement (projecting onto |0>) is treated as a measurement that reported incorrectly.
- measurement : M
The measurement whose outcome is flipped.
Instances For
Spacetime Faults #
A spacetime fault is a collection of space-faults and time-faults occurring at various locations and times during the procedure. We represent it as a pair of finsets: one of space-faults and one of time-faults.
The weight of a spacetime fault is the total number of individual faults (each single-qubit Pauli error counts as 1, each measurement error counts as 1).
- spaceFaults : Finset (SpaceFault Q T)
The collection of space-faults (Pauli errors on qubits at specific times).
The collection of time-faults (measurement errors).
Instances For
The weight of a spacetime fault: total number of individual faults.
Equations
- F.weight = F.spaceFaults.card + F.timeFaults.card
Instances For
The empty spacetime fault: no errors at all.
Instances For
A spacetime fault consisting of a single space-fault.
Instances For
A spacetime fault consisting of a single time-fault.
Instances For
Composition of spacetime faults via symmetric difference. Applying the same error twice cancels it; flipping an outcome twice restores it.
Equations
- F₁.compose F₂ = { spaceFaults := symmDiff F₁.spaceFaults F₂.spaceFaults, timeFaults := symmDiff F₁.timeFaults F₂.timeFaults }
Instances For
A spacetime fault is pure-space if it has no time-faults.
Equations
- F.isPureSpace = (F.timeFaults = ∅)
Instances For
A spacetime fault is pure-time if it has no space-faults.
Equations
- F.isPureTime = (F.spaceFaults = ∅)
Instances For
The number of space-faults in the spacetime fault.
Equations
- F.spaceWeight = F.spaceFaults.card
Instances For
The number of time-faults in the spacetime fault.
Equations
- F.timeWeight = F.timeFaults.card
Instances For
Basic Properties #
Composition Properties #
Space-only and time-only projections #
The space component of a spacetime fault (keeping only space-faults).
Equations
- F.spaceComponent = { spaceFaults := F.spaceFaults, timeFaults := ∅ }
Instances For
The time component of a spacetime fault (keeping only time-faults).
Equations
- F.timeComponent = { spaceFaults := ∅, timeFaults := F.timeFaults }
Instances For
Any spacetime fault decomposes into its space and time components.
Pauli operator associated to space-faults at a given time #
The space-faults of a spacetime fault restricted to a specific time.
Equations
- F.spaceFaultsAt t = {f ∈ F.spaceFaults | f.time = t}
Instances For
The composite Pauli error on qubit system Q from all space-faults at time t.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Convention: Initialization faults as time-faults #
The convention that initialization faults are treated as time-faults is captured by modeling each initialization as a measurement. An initialization fault (e.g., preparing |1> instead of |0>) is then a time-fault on that measurement, equivalent to perfect initialization followed by a Pauli X error (space-fault).
An initialization fault viewed as a time-fault on the initialization measurement.
Equations
- SpacetimeFault.initializationFault initMeasurement = SpacetimeFault.ofTimeFault { measurement := initMeasurement }
Instances For
The equivalent space-fault: perfect initialization followed by Pauli X error.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Weight bounds #
Fault classification predicates #
A spacetime fault is trivial (identity) if it has no faults at all.
Equations
- F.isTrivial = (F = SpacetimeFault.empty)
Instances For
The set of qubits affected by space-faults in a spacetime fault.
Equations
Instances For
The set of times at which space-faults occur.
Equations
Instances For
The set of measurements affected by time-faults.
Equations
Instances For
Measurement outcome with faults #
A time-fault flips the measurement outcome. In ZMod 2 encoding (0 -> +1, 1 -> -1), this means adding 1 to the ideal outcome.
The observed measurement outcome given an ideal outcome and a set of time-faults. If measurement m appears in the time-faults, the outcome is flipped.