Special | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | ALL
A |
---|
Abstract algebra | |||
---|---|---|---|
See 👇 attachment or the latest source on Overleaf, on the topics of groups and Galois fields (pending). | |||
Ancilla | ||||||
---|---|---|---|---|---|---|
An ancilla (system) is an auxiliary quantum-mechanical system [ETS18]. Think of ancilla as something extra that is used to achieve some goal [Pre18]. References
| ||||||
Axiomatic approach to quantum operations | ||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
In quantum information theory, the foundation of quantum cryptography, the concepts of general/generalised measurements and quantum channels are rooted in the concept of quantum operation [Pre18, Sec. 3.2.4].
While the time evolution of a closed quantum system can be expressed using a single unitary operator, the time evolution of an open quantum system is much less straightforward [HXK20], and this is where the theory of quantum operations come in. Here, the axiomatic approach in [NC10, Sec. 8.2.4] to defining quantum operations is discussed, starting with Definition 1. Definition 1: Quantum operation [NC10, p. 367]
A quantum operation is a map from the set of density operators for the input space to the set of density operators for the output space , that satisfies these axioms:
Justification of Axiom 1: In the absence of measurement, by itself completely describes the quantum operation, and Axiom 1 reduces to the requirement that ; in this case, the quantum operation is trace-preserving.
Justification of Axiom 2: Let , then where is a normalisation factor. By Bayes’ rule, Justification of Axiom 3: A quantum operation maps a density operator to a density operator — by itself or as part of a larger system — and density operators are always positive, so must be a completely positive map. Example 1 [Pre18, p. 19]
This example is meant to show that the transpose operation is not a completely positive map and hence not a quantum operation. We first observe that is positive if is positive because the quadratic form However, the transpose operation is not completely positive due to the following. Suppose system contains entangled subsystems and , and has state , ignoring the normalisation constant. Above, we applied the identity . Now consider the composite map , where is the transpose operation, which transforms to . Applying the composite map to , we get Applying the composite map to the above, we get back Therefore, if we represent the map with a square matrix, the matrix is involutory (square = identity), and it is trivial to show that the eigenvalues of an involutory matrix are (see below), implying the composite map is not positive. In other words, the transpose operation is not a completely positive map, and hence not a quantum operation. Axioms 1-3 lead to the following important theorem. Theorem 1 [NC10, Theorem 8.1]
In Theorem 1,
Example 2 [MM12, p. 176]
The freedom in the operator-sum representation is especially useful for studying quantum error correction. Theorem 2 [NC10, Theorem 8.3]
Watch lectures on Kraus representations by Artur Ekert, inventor of the E91 QKD protocol: References
| ||||||||||||||||||
B |
---|
BB84: Overview | ||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Quantum key distribution (QKD) is a method for generating and distributing symmetric cryptographic keys with information-theoretic security based on quantum information theory [ETS18]. A QKD protocol establishes a secret key between two parties — let us call them 👩 Alice and 🧔 Bob as per tradition — connected by 1️⃣ an insecure quantum channel and 2️⃣ an authenticated classical channel [Gra21, Sec. 3.1].
A QKD protocol typically proceeds in two phases [Wol21, Ch. 4]:
The security of QKD hinges on the principles of quantum mechanics, rather than the hardness of any computational problem, and hence does not get threatened by advances in computing technologies.
Theorem 1: No-cloning theorem [WZ82]
It is not possible to perfectly clone an unknown quantum state. The earliest QKD protocol is due to Bennett and Brassad [BB84] and is called BB84, named after the authors and the year it was proposed. QKD leverages physical mechanisms, so unavoidably we need to discuss the physical mechanisms that underlie/enable BB84, which are based primarily on the polarisation of photons [Wol21, Sec. 1.3.1]. PolarisationThe polarisation of photons specifies the geometrical orientation of the oscillation of its electromagnetic field.
We only consider linear polarisation here. For linear polarisation, we distinguish between two bases:
Consider the effect of polarisation filters depicted in Fig. 1:
Thus, measuring a diagonally polarised photon in the rectilinear basis, and similarly measuring a vertically/horizontally polarised photon in the diagonal basis, give a random result.
With knowledge of polarisation in mind, let us now discuss the quantum transmission phase of BB84, which involves encoding of classical bits into quantum states, communication over a quantum channel, and decoding of quantum states into classical bits. Quantum transmission phaseThis phase of the protocol involving 👩 Alice and 🧔 Bob goes like this [Wol21, Sec. 1.3.2]:
An illustration of the process above can be found Fig. 1. Since the polarisation state of each photon is a discrete variable, BB84 is an example of a discrete-variable quantum key distribution (DV-QKD) scheme. BB84 is also an example of a prepare-and-measure protocol, because of the preparation action of 👩 Alice and the measurement action of 🧔 Bob. Classical post-processing phaseThis phase of the protocol involves 👩 Alice and 🧔 Bob exchanging a sequence of classical information in the classical channel to transform their raw key pair into a shared secret key [Wol21, Sec. 1.3.3]:
Table 1 shows an example of an exchange between 👩 Alice and 🧔 Bob in the absence of eavesdropping. For discussion of physical realisations of BB84, follow this knowledge base entry. Performance and security evaluationIn terms of performance, a basic figure of merit of every QKD protocol is the secret key rate, i.e. the fraction of secure key bits produced per protocol round. For security, follow this overview of QKD security. References
| ||||||||||||||||||||||||||||
BB84: Physical realisations | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Acknowledgement: Andrew Edwards contributed some explanation.
Continuing from an overview of BB84, this entry discusses several ways in which BB84 can be realised. Fig. 1 shows an experimental setup built by IBM [NC10, Box 12.7], where
For the setup in Fig. 1,
Fig. 2 shows a minimalist block diagram for EDU-QCRY1, while Fig. 3 shows a photo of a physical setup realising the block diagram. Let us study the functions of the PBS in this context:
In recent years, satellite-based experiments on BB84 and extensions of BB84 (e.g., decoy-state BB84) had been conducted [LCPP22]. Compared to free space, polarisation is harder to preserve over commercial optical fibres [GK05, Fig. 11.7]. An alternative approach to polarisation is using an interferometer, such as a Mach-Zehnder interferometer; see Fig. 7 and [HIP+21, Sec. 3.2]. References
| ||||||||||||||
Bell states | ||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Consider the circuit below, where a Hadamard gate is connected to qubit and a controlled-NOT (CNOT) gate is connected to after the Hadamard gate: The Hadamard gate in Fig. 1 effects the transformations: and . The CNOT gate in Fig. 1 has its control qubit connected to the line, and its target qubit connected to the line. The unitary matrix representing the CNOT gate is If the input to the CNOT gate is the state , then the output is Thus, as an example, if the input is , then the Hadamard gate transforms it to , and the CNOT gate further transforms it to Table 1 is the truth table summarising the outputs corresponding to basis-state inputs.
The output states in Table 1 are called the Bell states or Einstein-Podolsky-Rosen (EPR) pairs [KLM07, p. 75; NC10, Sec. 1.3.6], and can be represented concisely as When the Bell states are used as an orthonormal basis, they are called the Bell basis [Wil17, pp. 91-93]. References
| ||||||||||||||||||
Bloch sphere | ||||||
---|---|---|---|---|---|---|
Consider the ket vector , where and are complex-valued probability amplitudes, and are the computational bases. We can express and in the exponential form [Wil17, (3.6)]: Thus, given , we can rewrite it in the physically equivalent form: 💡 Above, is used instead of because for visualisation using a Bloch sphere (more on this later), as ranges from 0 to , the values of and are confined within to ensure the qubit representation is unique [Wil17, p. 57]. So what is a Bloch sphere? Named after physicist Felix Bloch, a Bloch sphere is a unit-radius sphere for visualising a qubit relative to the computational basis states. As shown in Fig. 1,
As explained in Fig. 2, a sphere provides the necessary number of degrees of freedom to represent a ket vector. The equator of a Bloch sphere enables the representation of complementary bases. For example, in Fig. 3,
References
| ||||||
C |
---|
Coherent attack | |||
---|---|---|---|
The discussion here follows from the discussion of collective attacks. | |||
Coherent state | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
A coherent state is a special quantum state that a coherent laser ideally emits [Wil07, p. 19]. That 👆 does not say much, but there is no straightforward way to define “coherent state”, a concept introduced by Schrödinger [SZ97, p. 46]. Mathematically and in short, a coherent state is the eigenstate of the positive frequency part of the electric field operator [SZ97, p. 46], but this requires definition of the electric field operator, which in turn requires discussion of the quantization of electromagnetic fields. Classically, an electromagnetic field consists of waves with well-defined amplitude and phase, but in quantum mechanics, this is no longer true. More precisely, there are fluctuations in both the amplitude and phase of the field [SZ97, Ch. 2]. A coherent state is a state that has the same fluctuations of quadrature amplitudes as the vacuum state but which possibly has nonzero average quadrature amplitudes [Van06, Sec. 4.6.1]. References
| ||||||||||
Collective attack | ||||||||
---|---|---|---|---|---|---|---|---|
The discussion here follows from the discussion of individual attacks. In a collective attack [Sch10],
References
| ||||||||
Composite quantum systems | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Consider a multi-qubit system consisting of two qubits. There are four possible final states: ; and it makes sense to express the quantum state of this two-qubit system as the linear combination: where the normalisation rule still applies: An alternative representation of is where the subscript indicates the order of the system. Some authors call the set , and equivalently , the tensor base [Des09, p. 316]. Watch Microsoft Research’s presentation on “Quantum Computing for Computer Scientists”: In the vector representation, given two separated qubits: their collective state can be expressed using the Kronecker product (also called the matrix direct product and tensor product): Multiple shorthands exist for : 1️⃣ , 2️⃣ , 3️⃣ [Mer07, p. 6; NC10, Sec. 2.1.7]; we have used the third shorthand earlier. In general, we can compose the Hilbert space of a multi-qubit system using the vector space direct product (also called tensor direct product) of lower-dimensional Hilbert spaces [NC10, Sec. 2.1.7]:
Classical-quantum stateIn quantum cryptography, we often encounter states that are partially classical and partially quantum. A classical state (c-state for short) is a state defined by a density matrix that is diagonal in the standard basis of the -dimensional state space of , i.e., has the form: Suppose we prepare the following states for Alice and Bob: with probability 1/2, we prepare and with probability 1/2, we prepare . Then, the joint state is the so-called classical-quantum state (cq-state for short): In quantum-cryptographic convention,
Formally,
References
| ||||||||||||
E |
---|
Entropy | ||||||||
---|---|---|---|---|---|---|---|---|
The following introduces Shannon entropy before von Neumann entropy. Shannon entropyThe Shannon entropy of a random variable, say , measures the uncertainty of . Intuition:
The following definition of entropy, denoted by , measured in number of bits, can reflect the two extreme cases above [MC12, Definition 5.4]: where denotes the probability of taking on the th value. ⚠ Note: 1️⃣ ; 2️⃣ number of bits is discrete in practice but as a metric of comparison, we need entropy to be a continuous-valued metric. If has possible values, and has , the joint entropy of and is defined as [MC12, Definition 6.2]: The conditional entropy of given is defined as [MC12, (6.53)]: The chain rule for entropy states [MC12, p. 151; Gra21, (2.48)]: or more generally, Von Neumann entropyThe Shannon entropy measures the uncertainty associated with a classical probability distribution. Quantum states are described in a similar fashion, with density operators replacing probability distributions. The von Neumann entropy of a quantum state, , is defined as [NC10, Sec. 11.3]: where denotes the base-2 logarithm of and not the element-wise application of base-2 logarithm to . Watch an introduction to the matrix logarithm on YouTube: References
| ||||||||
F |
---|
Fidelity | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
This entry continues from discussion of trace distance. [Des09, (22.8)]. [Hay17, Sec. 3.1.2 and Sec. 8.2]. [Le 06, (7.20)]. [Dio11, (6.19)]. [NC10, (9.2)]. [BS98, Sec. II]. References
| ||||||||||||||
G |
---|
Galois field / finite field | |||
---|---|---|---|
See abstract algebra.
| |||
Group theory | |||
---|---|---|---|
See abstract algebra.
| |||
I |
---|
Idempotence | ||||||||
---|---|---|---|---|---|---|---|---|
A square matrix, , is idempotent if it is equal to its square, i.e., [Ber18, Definition 4.1.1]. Notable properties: References
| ||||||||
Individual attack | ||||||||
---|---|---|---|---|---|---|---|---|
The discussion here continues from the discussion of security of quantum key distribution (QKD). QKD is a method for generating and distributing symmetric cryptographic keys with information-theoretic security based on quantum information theory [ETS18]. In a QKD protocol,
😈 Eve’s attempt to discover the secret key can be classified into 1️⃣ individual attack (discussed below), 2️⃣ collective attack, and 3️⃣ coherent attack; in 🔼 increasing order of power given to Eve. Individual attacks are the simplest and most studied class of attacks. When conducting an individual attack, Eve interacts with each signal (i.e., quantum state) from Alice individually, and is restricted to the same interaction for all Alice’s signals [Sch10, Wol21].
Regardless, Eve has the freedom to choose which unitary operation she applies to a composite system (“composite” because more than one qubit is involved). Of interest is the amount of information about a single state that Eve gains with an attack. A standard measure of amount of information is mutual information. Perhaps the most intuitive example of an individual attack is the intercept and resend (I&R) attack [Sch10, Sec. 5.2.1]. References
| ||||||||
K |
---|
Kraus operator and Kraus representation | ||
---|---|---|
L |
---|
Linear operator | ||||
---|---|---|---|---|
Let and be vector spaces. The mapping is called a linear transformation if and only if for every choice of and scalar . When , is called a linear operator [DG09, p. 202]. References
| ||||
Lipschitzness, Lipschitz condition | ||||||
---|---|---|---|---|---|---|
Mathematical programming (theory-based optimisation methods as opposed to heuristics) works best with differentiable cost/loss functions. Mathematical programming also works with continuous loss functions [Byr15]. Differentiability implies continuity but the converse is not true, so continuity is a weaker condition than differentiability. For example, piecewise continuous functions are not differentiable at all points. Lipschitzness is a particular form of continuity. Strictly speaking, Lipschitzness is a form of uniform continuity: Definition 1: Lipschitzness [SSBD14, Definition 12.6]
It follows from the definition above that if the derivative of is everywhere bounded in absolute value by , then is -Lipschitz. References
| ||||||
M |
---|
Mutual information | ||||
---|---|---|---|---|
The following introduces classical mutual information before quantum mutual information. Classical mutual informationConsider the channel model of a transmission system in Fig. 1. Since the input takes a certain value at a certain probability, the input is a discrete random variable, which we denote by .
Suppose the sender transmits the th input symbol as at a probability of , this probability is called the prior probability of . “Prior probability” is often written in Latin as “a priori probability”. Suppose the receiver receives as the th output symbol, the probability of this output conditioned on the th input being is the likelihood of : . However in most cases, we are more interested in the posterior probability of : , i.e., what is the probability that the th input is given the th output is . “Posterior probability” is often written in Latin as “a posteriori probability”. The posterior probability helps us determine the amount of information that can be inferred about the input when the output takes a certain value. The information gain or uncertainty loss about input upon receiving output is the mutual information of and , denoted by [MC12, pp. 126-127].
Extending the result above from to and from to , we define the system/average mutual information of and , denoted by , as the information gain or uncertainty loss about random variable by observing random variable [MC12, Definition 6.7]: It is trivial to show that 1️⃣ , 2️⃣ , 3️⃣ iff and are independent. Omitting the derivation in [MC12, p. 129], where denotes the Shannon entropy of . Fig. 1 depicts the relation between mutual information and different entropies. Quantum mutual informationQuantum mutual information is mutual information where the entropy is von Neumann entropy instead of Shannon entropy. References
| ||||
P |
---|
Photon number splitting attack | ||||||
---|---|---|---|---|---|---|
References
| ||||||
Positive operator-valued measure (POVM) | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
The positive operator-valued measure (POVM) is a mathematical formalism/tool representing a measurement operation [NC10, Sec. 2.2.6; WN17, Sec. 1.5.1] that satisfies Postulate 3. Suppose a measurement described by measurement operators , where takes value from a finite set, is performed on a quantum system in the state , then the probability of outcome is given by Suppose we define then satisfies 1️⃣ the completeness relation and 2️⃣ is positive-semidefinite.
Example 1 [NC10, p. 92]
Suppose Alice gives Bob a qubit prepared in one of two states: Here is a measurement strategy for Bob to determine unequivocally whether he receives or . Define: Clearly satisfy the completeness relation, and by checking the definition of positive semidefiniteness, we can verify to be positive operators, so form a POVM. Let us now see how Bob can use to distinquish between and :
Using , Bob never mistakes for and vice versa, but Bob sometimes cannot determine which state he receives. Example 2 [WN17, Example 1.5.1]
Denote by an orthonormal basis. Define (a projector), and , then for each , where is a matrix with the basis vectors as columns. The linear independence of the basis vectors implies is nonsingular, and In other words, satisfies the completeness relation. When the POVM elements are projectors, such as in the preceding example, the POVM is called a projection-valued measure (PVM) [Hay17, p. 7]. Watch edX Lecture 1.6 “Generalized measurements” of Quantum Cryptography by CaltechDelftX QuCryptox. References
| ||||||||||||
Positive semidefiniteness and positive definiteness | ||||||
---|---|---|---|---|---|---|
For Hermitian matrix , the following statements are equivalent, and any one can serve as the definition of the positive definiteness of [Mey00, Sec. 7.6; Woe16, Sec. 7.4]:
Similarly, the following statements are equivalent, and any one can serve as the definition of the positive semidefiniteness of : References
| ||||||
Projective measurement | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
The discussion here follows from the discussion of the positive operator-valued measure (POVM). The POVM cannot be used to determine the post-measurement state because the post-measurement state may not be pure [WN17, p. 13]. Instead, a Kraus operator representation of the POVM, known as projective measurement, is necessary to specify the post-measurement state. Projective measurements are repeatable, and the outcome observed as a result of projective measurement is deterministic [MM12, Sec. 2.7], as explained below.
Projective measurements are often discussed in terms of an observable, which is defined as a Hermitian operator acting on the state space of a system. Every observable has a spectral decomposition [Mey00, p. 517]: where are the eigenvalues of , and are so-called spectral/orthogonal projectors onto the null space of . More precisely, 1️⃣ are Hermitian and idempotent, 2️⃣ , and 3️⃣ are mutually orthogonal. Clearly, are valid POVM elements. A (non-unique) Kraus operator representation or Kraus decomposition of is defined as [WN17, Definition 1.5.2]:
With Kraus operators, we can now define projective measurement: Definition 1: Projective measurement [WN17, Definition 1.5.4; MM12, p. 153]
A projective measurement is given by a set of orthogonal projectors such that . By default, the Kraus operator is chosen to be . The probability of observing measurement outcome given initial state is The post-measurement state is
Since the default Kraus operators are the same as the orthogonal projectors, most authors skip discussion of Kraus operators altogether 🤷♂️. Measuring the observable is equivalent to performing a projective measurement with respect to the decomposition , where the measurement outcome corresponds to eigenvalue . Example 1 [KLM07, Example 3.4.1]
Consider the Pauli observable , i.e., the Pauli-Z operator/matrix: which does not change an input of , but flips to (which is equivalent to with a phase change). The Pauli-Z operator acts as a NOT operator in the Hadamard basis. has eigenpairs and , where the eigenvectors and are also called eigenstates [MM12, p. 338]. Thus, has spectral decomposition: with orthogonal projectors Interpreting the above, a projective measurement of is a measurement in the standard basis with eigenvalue corresponding to final state and eigenvalue corresponding to final state . Example 2 [WN17, Example 1.5.3]
Given a two-qubit state , suppose we want to measure the parity of the two qubits in the standard basis. One method is to measure in the (-dimensional) standard basis, obtain two classical bits, and take their parity. In this case, the probability of obtaining outcome “even” is and the post-measurement state is Another method is to measure the parity using projective measurement which directly projects onto the relevant subspaces, without measuring the qubits individually. Define projectors: which we can quickly verify to be orthogonal and conformant with the completeness relation: In this case, the probability of getting outcome “even” is which is the same as before; and the post-measurement state is The preceding two methods produce different outcomes on the EPR state , or in density-matrix representation, By measurement in the standard basis, the probability of getting outcome “even” is while the post-measurement state is By projective measurement, is the same as before, but the post-measurement state is different: Thus, projective measurement with does not change the EPR state. This is one of the main advantages of using projective measurement as opposed to basis measurement: the former enables simple computation (e.g., parity) on multi-qubit states without fully “destroying” the state, as basis measurement does. In Example 2, the observation that the projective measurement preserves the quantum state is not a coincidence. In fact, building on Definition 1, applying measurement to , i.e., applying to twice gives us [MM12, p. 153]: References
| ||||||||||||||||
Projector | ||||
---|---|---|---|---|
A matrix is a projector if it is 1️⃣ Hermitian and 2️⃣ idempotent [Ber18, Definition 4.1.1]. References
| ||||
Pure state, mixed state, density operator/matrix | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
As defined in Practical 1 of COMP 5074, a qubit is a quantum state with two possible outcomes. Mathematically, a qubit is a two-dimensional Hilbert space. A pure state is the quantum state of a qubit that we can precisely define at any point in time [Qis21, Sec. 1].
In non-ideal conditions, a qubit can assume a certain quantum state at a certain probability, and other quantum states at other probabilities, regardless of measurements.
A direct but rather clunky representation of a mixed state is the set of pairs: where each pair means the pure state appears at probability , for . Note . A more compact and useful representation of a mixed state can be obtained from the type of Hilbert-space operators called density operators [KLM07, Sec. 3.5.1].
If the explanation above is hard to grasp, see if edX Lecture 1.2 “The density matrix” of Quantum Cryptography by CaltechDelftX QuCryptox helps. When applied with unitary operator , state evolves into . Accordingly, the density matrix of the mixed state evolves into Our adventure with the trace operator continues because an important application of the trace operator is determining the probability of getting result when applying measurement operator to mixed state . By definition, the probability of getting result given initial state is [NC10, p. 99]: The last equality is due to Eq. (1). By the law of total probability, Omitting the derivation in [NC10, pp. 99-100], the density matrix of the system after making measurement and obtaining result is: By the law of total probability, implying which is called the completeness equation. ⚠ Note is unitary if and only if has one value. Postulates of quantum mechanicsUsing the language of the density operator, we can phrase the fundamental postulates of quantum mechanics as [NC10, pp. 98-102; Gra21, p. 11-12]: Postulate 1: Associated with any isolated physical system is a complex-valued Hilbert space known as the state space of the system. The system is completely described by its density operator, which is a positive operator with trace one, acting on the state space of the system. If a quantum system is in the state (in density-matrix, not ket notation) with probability , then the density operator for the system is . Postulate 2: The evolution of a closed quantum system is described by a unitary transformation. That is, the state of the system at time is related to the state of the system at time by a unitary operator , which depends only on the times and : . Postulate 3 (aka Measurement Postulate): Quantum measurements are described by a collection of measurement operators. These are operators acting on the state space of the system being measured. The index refers to the measurement outcomes that may occur in the experiment. If the state of the quantum system is immediately before the measurement then the probability that result occurs is given by and the state of the system after the measurement is The measurement operators satisfy the completeness equation: Postulate 4: The state space of a composite physical system is the tensor product of the state spaces of the component physical systems. Moreover, if we have systems numbered 1 through , and system number is prepared in the state , then the joint state of the total system is . Postulate 3 defines general/generalised measurement [NC10, Box 2.5], which consists of 1️⃣ a rule describing the probabilities of different measurement outcomes, and 2️⃣ a rule describing the post-measurement state.
References
| ||||||||||||||
Q |
---|
Qiskit: setup | |||
---|---|---|---|
This entry continues from installation of QuTiP.
Qiskit comes with a super-quick introduction to Python and Jupyter notebooks. Visual Studio Code is a popular alternative to a web browser for hosting Jupyter notebooks, but to minimise the number of software components that can go wrong, we shall stick with a web browser (for which Mozilla Firefox is recommended). | |||
Quantum bit error rate (QBER) | ||||
---|---|---|---|---|
The quantum bit error rate (QBER) is the error rate affecting two bitstrings obtained as a result of the same quantum measurement performed on two different sets of quantum systems [Gra21, p. xvi]. See BB84 to put this definition in context. References
| ||||
Quantum supremacy | ||
---|---|---|
Qubit: physical realisation | ||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Creating a low voltage to represent a logical “0” and a high voltage to represent a logical “1” is straightforward. 🤷♂️ Creating a superposition of low and high voltages however is not. A quantum computer is commonly envisioned to be a machine that exploits the full complexity of a many-particle quantum wavefunction to solve computational problems [LJL+10].
For the construction of quantum computers, laser serves as an inspiration because it is quantum mechanics that enables laser waves to be generated in phase [LJL+10]. Just as there are many possible materials for lasers (e.g., crystals, organic dye molecules, semiconductors, free electrons), there are many materials under consideration for quantum computers; see [LJL+10] and [MM12, Ch. 6]. Quantum bits are often imagined to be constructed from the smallest form of matter, e.g., an isolated atom, through ion traps and optical lattices, but they can also be made in components far larger than consumer electronics, e.g., a superconducting system [LJL+10]. Below we discuss three main technologies [LLSK22, Sec. 6]: 1️⃣ trapped ions, 2️⃣ photonics, and 3️⃣ superconducting qubits. Trapped ions: Main idea is to use the two different internal states of a trapped atomic ion as a two-level system (i.e., qubit) [LLSK22, Sec. 6.2]. An ion trap uses electromagnetic fields and laser cooling to control the spatial position of an ion in vacuum and reduce the temperature of the ion [LLSK22, Sec. 6.2; BCSH21, Sec. 2]. Watch an introduction to the ion trap: Lasers or microwaves are used to control the internal states of an ion [BCSH21, Figure 1]. The internal control plus the Coulomb repulsion between ions combine to form conditional logic gates [BCSH21, Figure 1]. 👍: State preparation, qubit measurement, single-qubit and two-qubit gates can be performed with fidelities (> 99%) higher than what is required for quantum error correction [LLSK22, Sec. 6.2]. 👎: A large array of bulk optical components are necessary and these are difficult to address and measure, challenging scalability [LLSK22, Sec. 6.2]. Trapped-ion quantum computers (e.g., IonQ) are enjoying a reasonable level of commercial success [LLSK22, Sec. 6.2]. Photonics: Photonics has always been a prominent candidate for realising qubits [LLSK22]. For generating qubits, photonics offers the following advantages [SP19, PAB+20, LLSK22]:
Superconducting qubits: They are currently the leading contenders in the race for large-scale quantum computing [LLSK22]. Superconducting qubits are the technology big-tech companies like Google and IBM have been focusing on. In 2019, a large research team consisting of Google and multiple American and European universities demonstrated “quantum supremacy” on a programmable superconducting quantum processor called “Sycamore”, which consists of a two-dimensional array of 54 transmon qubits [AAB+19]:
References
| ||||||||||||||||||
QuTiP (Quantum Toolbox in Python): setup | |||
---|---|---|---|
There is no shortage of quantum computing frameworks/toolkits, including Google’s Cirq, Rigetti’s Forest SDK (including pyQuil), Microsoft’s Q#-based Quantum Development Kit, IBM’s Qiskit, Quipper, QuTiP, ETHZ’s Silq, Cambridge Quantum Computing’s tket. We shall use Qiskit because 1️⃣ it is among the most well-established, and 2️⃣ it comes with rich learning resources.
We shall also use QuTiP, because while Qiskit is popular for quantum computing, QuTiP offers more features for quantum-dynamical simulations.
Watch an introduction to QuTiP: Using Qiskit and QuTiP means we are using Python but there are so many resources for learning Python it should not be an issue for a Bachelor/Master student to pick it up along the way. Even though the relevant computer pools will provide you with the required software, you need to install the software on your own computer(s) anyway, because you will not be able to finish any of the practicals within the allocated time. The allocated time is only for you to get enough supervision so that you can complete the remainder of the practical on your own. Our strategy here is to first install QuTiP, then Qiskit. The operating system of choice for many computer scientists and cryptograhers is Linux, but the setup guide here is only applicable to Windows, because UniSA computer pools have only Windows. Follow the instructions below (derived from the official instructions) to set up QuTiP:
Above, the backtick (`) is the line continuation character in PowerShell. Among the packages just installed, Official Conda documentation discourages using pip and conda together, but we need pip for Qiskit later. | |||
R |
---|
Reduced density operator/matrix and partial trace | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
The reduced density operator is an application of the density operator. The reduced density operator is so useful it is indispensable in the analysis of composite quantum systems [NC10, Sec. 2.4.3]. Consider a composite system, with density matrix , consisting of Alice’s system and Bob’s system, there is often a need to express the state of Alice’s or Bob’s system in terms of . The operation that takes us from to (denoting Alice’s density matrix) or (denoting Bob’s density matrix) is called the reduced density operator. The reduced density operator for Alice’s system is defined by where is a map of operators known as the partial trace over Bob’s system. Suppose and are any two vectors in Alice’s state space, and furthermore, and are any two vectors in Bob’s state space, then the partial trace is defined by the operation [NC10, (2.178)]: Above, the notation is equivalent to To see why the definition above makes sense, suppose a quantum system is in the product state , where and are the density matrices for subsystems and respectively, then since the trace of any density matrix is 1. Similarly, In general, if where is an orthonormal basis of , and is an orthonormal basis of , then the partial trace over is [WN17, Definition 1.6.1]: For a quick summary of discussion up to this point, watch edX Lecture 1.7 “The partial trace” of Quantum Cryptography by CaltechDelftX QuCryptox. Consider the pure (entangled) Bell state: . The system comprises 1️⃣ single-qubit subsystem with basis vectors and , and 2️⃣ single-qubit subsystem with basis vectors and . This system is entangled (i.e., not separable) because , but using the reduced density operator, we can find a full description for subsystem and for subsystem . The density matrix for the system is: Applying Eq. (1), the reduced density operator for subsystem is: Since , both and are mixed states. 🤔 How do we reconcile the preceding observation ☝ with the fact that is a pure state?
The strange property, that the joint state of a system can be pure (completely known) yet a subsystem be in mixed states, is a hallmark of quantum entanglement.
⚠ Caution [KLM07, Exercise 3.5.5]
The partial trace contains all the relevant information about subsystem if subsystem is discarded. Similarly, contains all the relevant information about subsystem if subsystem is discarded. These local descriptions do not in general contain enough information to reconstruct the state of the whole system. Expressing a bipartite vector in the Schmidt basis makes it much easier to compute the partial trace of either subsystem. For this reason, let us discuss Schmidt decomposition. References
| ||||||||||||
S |
---|
Schatten norm | ||||
---|---|---|---|---|
For , suppose its singular values are . Then, the Schatten -norm of , where , is defined as [Ber09, Proposition 9.2.3]: When , we have the Schatten 1-norm, which is also called the trace norm or nuclear norm [Ber09, p. 549]: When , we have the Frobenius norm: References
| ||||
Schmidt decomposition | ||||
---|---|---|---|---|
References
| ||||
Secret key rate | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
The secret key rate is the fraction of secure key bits produced per protocol round, where a round is the transmission of a quantum state through the quantum channel [Gra21, p. 38]. The secret key rate generally depends on the total number of rounds performed. The asymptotic secret key rate (often just asymptotic key rate) is the secret key rate when is assumed to simplify analysis.
Assuming is not realistic, and the security of a QKD protocol has to be analysed assuming a finite and generally finite resources. Analysis of the secret key rate and associated security properties of a QKD protocol is called finite-key analysis [TLGR12], to be covered in the future. References
| ||||||||||
Security of quantum key distribution (QKD): overview | ||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Methods for analysing the security of quantum key distribution (QKD) schemes/protocols are still being developed [PAB+20, PR22].
The security of a QKD scheme is typically analysed in terms of the level of success of 1️⃣ individual attacks, 2️⃣ collective attacks and 3️⃣ coherent attacks; in 🔼 increasing order of power given to the adversary [Wol21, Sec. 5.3.1].
The security of a QKD scheme is also often analysed in terms of composability, short for universally composable security.
TODO: Trace distance criterion [PR22], security = correctness + secrecy [Gra21]. Asymptotic vs finite-key security analysis. Security evaluation of practical QKD implementations involves evaluating the level of success of “quantum hacking” (i.e., side-channel attacks on QKD). References
| ||||||||||||||||||
Separable vs entangled | ||||
---|---|---|---|---|
The joint state, , of two quantum systems and is separable [WN17, Definition 2.1.1] if there exists a probability distribution , and sets of density matrices and such that If such an expression for does not exist, is entangled [WN17, Definition 2.1.1]. Specifically, if is a pure state, then is separable if and only there exists and such that Example 1: [WN17, Example 2.1.3]
Example 2: [WN17, Example 2.1.2]
Example 3: [WN17, Example 2.1.1]
Example 4: [WN17, Example 2.1.4]
This example is meant to highlight the difference between the following two states: where . is separable whereas is not. Consider the outcomes of measuring subsystem in and in the standard basis and in the Hadamard basis. Measuring subsystem of in the Hadamard basis: Define the measurement operators to be and , where and . Clearly, and . Using projective measurement, the post-measurement state conditioned on measurement outcome is Let us work out the numerator and the denominator separately, starting with the numerator: The preceding equality follows from these identities, which you will derive in the practical: The denominator is: The preceding computation is tedious but straightforward given the right tool (e.g., NumPy). Combining the results for the numerator and denominator, we get Thus, upon measuring a on subsystem , subsystem can be in either or at equal probabilities; we say the reduced state on is maximally mixed. Measuring subsystem of in the Hadamard basis: The pair of and can be re-expressed as and , because ‘’ is orthogonal to ‘’ just as ‘’ is orthogonal to ‘’. The preceding statement implies when is measured on subsystem , subsystem is in state as well. References
| ||||
Singular value | ||||||
---|---|---|---|---|---|---|
Singular values, like eigenvalues, are an intrinsic property of a matrix. Unsurprisingly, they can be defined in terms of eigenvalues: Definition 1: Singular value [Ber09, Definition 5.6.1]
Singular values can also be defined through the operation called singular value decomposition: Definition 2: Singular value decomposition and singular values [Hog13, Sec. 5.6]
References
| ||||||
Spectral theorem and spectral decomposition | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Spectral theorem is one of the fundamental theorems of linear algebra [ABH09, Sec. 3.6]. Based on the spectral theorem, spectral decomposition is an essential tool in quantum theory [NC10, Box 2.2]. Multiple equivalent interpretations of the spectral theorem exist, e.g., [ABH09, Theorems 3.6.4 and 3.6.12]. The interpretation in Theorem 1 directly defines spectral decomposition, and is hence also called the spectral decomposition theorem. Theorem 1: Spectral (decomposition) theorem [Hol13, Theorem 8.23; Zha11, Theorem 3.4; KLM07, Theorem 2.4.3]
Any normal operator, , has an outer product representation [KLM07, Sec. 2.4; Mey00, p. 517]: where The outer products are projectors that satisfy
References
| ||||||||||||||
SymPy | ||
---|---|---|
SymPy is a Python library for symbolic computing. In symbolic computing, we reason with symbols rather than numeric values. When running SymPy in Google Colab, make sure you are using a WebKit-based browser such as Chrome or Edge. The first thing to do when using SymPy is creating symbols. There are seven ways to create a symbol. Right from the beginning, it is crucial to know how to make assumptions in SymPy.
| ||
T |
---|
Time evolution of quantum systems | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
At the highest level, the time-evolution postulate of quantum theory [KLM07, p. 44] states
The time-evolution of the state of a closed quantum system is described by a unitary operator. That is, for any evolution of the closed system, there exists a unitary operator such that if the initial state of the system is , then after the evolution, the state of the system will be .
But how do we arrive at the understanding of the role of the unitary operator? Consider the evolution of one quantum state: to another quantum state: where are basis state vectors. Suppose there exists a linear operator that captures this evolution: such that . Besides linearity and conservation of overlaps (overlap of a vector with itself = norm squared), there are other properties that must satisfy. Define as a time-dependent map of one quantum state to another: , then the additional properties that must satisfy are [Hir04, Sec. 8.3.1]: To satisfy 1️⃣ linearity, 2️⃣ conservation of overlaps and 3️⃣ decomposability, must be unitary; see proof in [Hir04, Lemma 8.3.1]. Think of a unitary operator as a matrix transformation using a unitary matrix. A unitary matrix is a matrix whose Hermitian conjugate / Hermitian adjoint / conjugate transpose is also its inverse: where is the identity matrix of the appropriate dimensions. About the notation, many quantum physicists prefer to use 1️⃣ instead of to denote Hermitian conjugate, and 2️⃣ instead of to denote the identity matrix. The unitarity of quantum evolution implies reversibility since . Unlike classical logic gates, quantum gates are governed by unitary matrices and are thus reversible. However, measurements are not reversible; this is known as the measurement paradox of quantum physics [Hir04, p. 23]. A unitary operator is a normal operator because it commutes with its Hermitian conjugate, i.e., [KLM07, Definition 2.4.1]. 👈 This property enables the spectral decomposition of a unitary operator. To additionally satisfy continuity, Stone’s theorem [Par92, Sec. I.13] necessitates the existence of a Hermitian / self-adjoint operator such that where is called the (quantum) Hamiltonian or Hamilton operator representing the total energy of the closed quantum system [Hir04, Theorem 8.3.1]. The exponential in Eq. (1) is a matrix exponential: More info about the matrix exponential is available in the knowledge base entry on state-space equations. Sometimes, Eq. (1) is written as , where is the Planck’s constant whose value must be experimentally determined [NC10, p. 82]. which is usually written in the following form: The linear differential equation above is called the general version of the time-dependent Schrödinger equation [Wei15, p. 82], which is sometimes called the abstract Schrödinger equation [Hir04, p. 131]. Summarising the discussion so far, the time-evolution postulate can be rephrased in more precise mathematical terms [NC10, Postulate 2']: References
| ||||||||||||||
Trace distance | ||||||||
---|---|---|---|---|---|---|---|---|
Measurement of information is crucial to cybersecurity. One of these measures is distance measure between two quantum states.
Distance measures are defined in a way that makes sense to the analysis they are applied to, hence more than one distance measure exist in the literature, but two of these measures are in particularly wide use, namely trace distance and fidelity. The focus here is trace distance, which for probability density functions and index set is defined to be [NC10, Eq. (9.1)]: Trace distance is also called distance and Kolmogorov distance. Trace distance satisfies the mathematical definition of metric, because it satisfies [NC10, p. 400]: Extending the earlier definition to quantum states, the trace distance between density matrices and is defined as [MM12, Sec. 3.11]: If References
| ||||||||
Trace norm | ||||||
---|---|---|---|---|---|---|
The trace norm is an example of a unitarily invariant norm and is equivalent to the Schatten 1-norm. References
| ||||||
Trace operator | ||||||
---|---|---|---|---|---|---|
The trace of a square matrix , denoted by tr or tr, is a linear map that maps the matrix to a complex number, and is specifically the sum of the diagonal elements of the matrix. Obvious properties: Useful property involving brakets:
References
| ||||||
U |
---|
Unitarily invariant norm | ||||
---|---|---|---|---|
A vector norm on is unitarily invariant if for any and unitary matrices of appropriate dimensions and , we have [Hog13, Sec. 24.3]. References
| ||||