Research Article  Open Access
Anuradha Mahasinghe, Sachiththa Bandaranayake, Kaushika De Silva, "Solovay–Kitaev Approximations of Special Orthogonal Matrices", Advances in Mathematical Physics, vol. 2020, Article ID 2530609, 7 pages, 2020. https://doi.org/10.1155/2020/2530609
Solovay–Kitaev Approximations of Special Orthogonal Matrices
Abstract
The circuitgate framework of quantum computing relies on the fact that an arbitrary quantum gate in the form of a unitary matrix of unit determinant can be approximated to a desired accuracy by a fairly short sequence of basic gates, of which the exact bounds are provided by the Solovay–Kitaev theorem. In this work, we show that a version of this theorem is applicable to orthogonal matrices with unit determinant as well, indicating the possibility of using orthogonal matrices for efficient computation. We further develop a version of the Solovay–Kitaev algorithm and discuss the computational experience.
1. Introduction
A computer program in the context of classical computing is an ordered list of instructions, expressible in terms of elementary operations, readily convertible to the machine language of a classical computer. A quantum program in quantum computing could be described analogously. According to the circuitgate framework of quantum computing, a quantum algorithm consists of quantum gates acting on quantum states (qubits) where measuring devices are applied at appropriate instances to collapse the wavefunction. Based on the Heisenberg–Born interpretation of quantum mechanics, this circuitgate framework has achieved significant progress up to date, as the pioneering model of quantum computing. Also, it is proven to be polynomially equivalent to other quantum computational frameworks. Accordingly, a quantum program can be regarded as the application of several unitary matrices, together with measurements at certain instances.
In order to implement unitary operations, basic quantum gates such as Pauli gates, Hadamard gate, and phase gate are available in the circuitgate framework, in analogy with basic gates in classical computing. It is quite natural to ask how many basic gates are needed to implement an arbitrary unitary operation in a quantum circuit. The remarkable contributions in this regard made independently by Solovay [1] and Kitaev [2] answered this question, resulting in what is known today as the Solovay–Kitaev theorem. This theorem states that it is possible to approximate any unitary with unit determinant by a product of physically realizable unitaries (which appear as basic gates) to an arbitrary accuracy [3, 4]. Recall the other quantum computational frameworks such as quantum walks, quantum Turing machines, and adiabatic computing were proven to be polynomially equivalent to the circuitgate framework [5–7], the Solovay–Kitaev theorem is widely regarded the theoretical proof for the supremacy of quantum computers. In addition, the number of elementary gates needed to implement an arbitrary unitary provides an indicator of the capacity and limitations of quantum computers.
This reveals an interesting aspect of unconventional models of computing. That is, any computational model with similar speed and limitations would be computationally equivalent to quantum computing. If physically realizable, such a model would have the same advantages and limitations as quantum computing. Though little attention has been paid to this subject in the past, several interesting works have investigated the possibility of having such models. A pioneering work was done by Aerts and Czachor in 2007, proposing geometric algebras instead of unitary matrices [8]. The authors named this model cartoon computing and proved its equivalence to quantum computing, demonstrating a simulation of the DeutschJozsa algorithm. A later work investigated entities in cartoon computing equivalent to elementary gates in quantum computing [9]. In 2008, Fernandez and Schneeberger proposed quaternionic computing, in which the possibility of adopting quaternions instead of unitary matrices was proven [10]. In order to show the equivalence to quantum computing, the authors have used the Bernstein–Vazirani theorem in quantum Turing machine framework. In [11], Graydon explored quaternionic quantum processes with respect to standard quantum information theory. Thus, it is an interesting question to ask what other algebraic structures would show similar behaviour, if employed as a computational model.
On the other hand, the progress achieved in threelevel quantum systems is noteworthy [12–14]. Instead of qubits with states and in standard circuitgate framework, qutrits having three basis states , , and are used in these systems. Analogous to the singlequbit quantum gates in the form of matrices in the special (unit determinant) unitary group , the singlequtrit gates are special unitaries or the elements in the group [15–17]. Though the realization of gates has been the topic of interest for several previous works [18–20], computational capacity or theoretical bounds on computing of a threelevel quantum system have not been paid the deserved attention. Neither Solovay–Kitaev type approximations were investigated for threelevel systems. Nevertheless, a recent work emphasized the significance of the subgroup of for qutritbased quantum computation, showing that any state of a qutrit could be obtained from a oneparameter family of states through the action of [21]. In this regard, one should not ignore the remarkable relationship between the groups and . This motivates us to check whether the Solovay–Kitaev theorem is extendable to and possible to achieve the quantum speedup in threelevel quantum systems, when equipped with orthogonal operators. In addition to that, once this question is resolved, one may know exactly whether the orthogonal matrices also provide an algebraic structure suitable for efficient computation, such as geometric algebras or quaternions.
In this paper, we show that the question is answered positively. That is, the orthogonal matrices play a role in threelevel quantum systems, equivalent to what the unitaries play in standard quantum circuit framework. More precisely, we show that a version of the Solovay–Kitaev theorem is applicable to orthogonal matrices with unit determinant. Thus, we indicate the possibility of theoretically replacing the unitaries in quantum computing by orthogonals and qubits by qutrits. Using Cornwell’s twotoone homomorphic map from the special unitary group to special orthogonal group [22], we prove the possibility of approximating any orthogonal with unit determinant by a product of elementary orthogonals of unit determinants to an arbitrary accuracy . We further discuss how to find the sequence of appropriate elementary orthogonals, providing a version of Solovay–Kitaev algorithm in .
The remainder of the paper is organized as follows. In Section 2, our version of the Solovay–Kitaev theorem for is proven. An approximation scheme for unitdeterminant orthogonal matrices in accordance with this theorem and the standard Solovay–Kitaev algorithm is presented in Section 3. Computational experience is discussed in Section 4, and we discuss the implications of our work in Section 5 with several remarks on potential future works.
2. Solovay–Kitaev Theorem in
2.1. Solovay–Kitaev Theorem
Recall the computational power in the circuitgate framework is guaranteed by the Solovay–Kitaev theorem; its main focus is on approximating a unitary matrix by basic quantum gates. A set of possible basic gates is referred to as an instruction set in the context of this theorem. Considering single qubit unitary gates, an instruction set is a finite subset of such that contains its own inverse and is dense in . For example, the set of gates makes an instruction set for , where and denote, respectively, Hadamard and phase gates in the circuitgate framework. The set of all strings that can be made from without using more than elements is denoted by . Now, the Solovay–Kitaev theorem for a qubit system can be stated as follows.
Theorem 1 (Solovay–Kitaev). Let be an instruction set in . Then, for any , provides an –net for where .
The proof of this theorem is highly constructive, and the algorithmic steps of finding the elements in the instruction set that approximate a given element in can be found from its proof. A comprehensive version of proof can be found in [4]. An algorithmic version of the theorem with a procedure for finding those elements can be found in [23]. We now explore how a version of this theorem can be adapted to . Primary motivation for this is the distance relations of the two groups, preserved by a homomorphism from onto .
2.2. Distance Relations
The twotoone homomorphic mapping from onto known today as Cornwell’s mapping is expressible in several ways [22], from which we adopt the following in [24]. An element in is expressible as where and and its image is given by
In order to measure the distances, as in the proof of the standard Solovay–Kitaev theorem, we too use the metric induced by the trace norm for consistency. It is customary to use the operator norm in quantum computation according to the matrix formulation of quantum mechanics. However, the standard proof of the Solovay–Kitaev theorem uses trace norm, as it helps to make the proof more comprehensive by incorporating a special property of the trace norm at some point. Since our intention is finding an analogous version in , it is more appropriate to consider the trace norm for matrices in as well.
Lemma 2 illustrates how the mapping preserves distances to an order in with respect to trace norm.
Lemma 2. For any two , if , then .
Proof. Due to unitary invariance of the trace norm, it suffices to show that whenever . We use the fact that any element in can be expressed as in equation (1) and the mapping given by equation (2). Then, from which we derive
Supposing , it is not difficult to see that the left side of equation (4) is bounded by as follows.
. Accordingly, and . Also, . Therefore, . Similarly, . Substituting these in equation (4), . Therefore, .
2.3. Instruction Sets in
In the context of single qubit unitary gates, an instruction set is a finite subset of such that contains its own inverse and is dense in . It is possible to adopt the same definition for instruction sets in . Interestingly, the image of an instruction set in under the homomorphism becomes an instruction set in . Lemmata 3 and 4 prove this claim.
Lemma 3. Let and be metric spaces, and let be a dense subset of . If is continuous and surjective, then is dense in .
Proof. Let . Then, and since is continuous, is closed. This implies . On the other hand, since is dense in , . Thus, , and therefore, .
Lemma 4. If is an instruction set in , then is an instruction set in .
Proof. Let be an instruction set in . Then, must contain its own inverse and is finite as is a homomorphism. Since is continuous and is dense in , by Lemma 3 is dense in . Clearly, since is a homomorphism, we have . Therefore, is an instruction set in .
2.4. Solovay–Kitaev Theorem in
With the results we derived above, it is now possible to establish a version of the Solovay–Kitaev theorem for .
Theorem 5. Let be an instruction set in . Then, is an instruction set for such that for any , provides an –net for where .
Proof. From Lemma 4, is an instruction set. Let . Then, there exists some such that . The Solovay–Kitaev theorem guarantees the existence of such that such that . By Lemma 2, . Since is a homomorphism, . That is, , where .
3. Approximations in
Now we describe how an arbitrary unitdeterminant orthogonal matrix can be approximated by , where is an instruction set in . Recall the proof of the Solovay–Kitaev theorem is highly constructive; it provides essential ingredients for finding the sequence of elements from the instruction set approximating the given unitary to a given accuracy . As implied by Theorem 5, our algorithmic version for too is based on the steps in finding those elements as in the proof of the original theorem.
For completion, we first describe the algorithm for finding the approximations in . We follow the procedure given by Dawson and Nielsen [23] in this regard.
3.1. Solovay–Kitaev Algorithm
As described in [23], the Solovay–Kitaev algorithm is explainable using the following lemma.
Lemma 6 [23]. Suppose and are unitaries such that , and also . Then,
The algorithm in can be expressed in pseudocode as follows.

The algorithm is a function which takes two inputs: is an arbitrary element in which we desire to approximate by , and a nonnegative integer which controls the accuracy of the approximation. This function returns sequence elements from an instruction set in which approximates to an accuracy of , a strictly decreasing function of . The Solovay–Kitaev algorithm is recursive and the recursion terminates when .
In this step, we find an approximation to . To find such an approximation, we have to assure that we have constructed an net: a set containing elements from such that for any unitary matrix we can find an approximation from it. Since is a constant and is dense in , we can build a gate net by enumerating and sorting a large number of elements from for sufficiently large but fixed positive integer and creating a search algorithm to find the closed approximation. If , then we find an approximation to :
If is an approximation to , then by the unitary invariance of the norm,
Thus, finding an approximation to with allows us to find an improved approximation (i.e., ) to . To find such an approximation, first we decompose , where are unitaries with , where is positive constant:
This decomposition is known as the balance group commutator. To find such a decomposition, we use the fact that any arbitrary unitary can be represented as a rotation in the Bloch sphere. If is a rotation by an angle about some axis on the Bloch sphere, consider satisfying
Then, if is a rotation by about the axis and is a rotation by about the axis, on the bloch sphere, then is conjugate to (i.e., ) for some unitary . Since and are unitary matrices, they are diagonalizable; moreover, they have the same eigenvalues. Thus, by diagonalizing and , we find a diagonal matrix and two unitary matrices and such that
Now, letting , we have and satisfying
Also, for sufficiently small , and satisfy for some positive constant .
Now, we find approximations to both and :
By replacing by and by in Lemma 6, the group commutator of and turns out to be a approximation to for some positive constant . Now, if , then . Hence, provides an improved approximation for . Accordingly, the value of is determined by this constant ; i.e., for this construction to guarantee that , the value of must be strictly less than (i.e., ). This algorithm concludes by returning the sequences of elements in that approximate the group commutator as well as .
3.2. Solovay–Kitaev Algorithm in
In light of the algorithmic steps described, now it is possible to provide the algorithmic version for as follows.

This algorithm is a function which takes two inputs: : an arbitrary element in which we intend to approximate and : a nonnegative integer which controls the accuracy of the approximation. This function returns a sequence of elements from an instruction set , where is an instruction set in , which approximates to an accuracy of , where is a decreasing function of , i.e, as .
In this step, we find such that , where is the homomorphic mapping from to , so that we can find a Solovay–Kitaev approximation for in ( is the Solovay–Kitaev function in ):
Supposing that for a given depth the SK function approximates any unitary matrix to accuracy , we find a approximation to . Next, we find :
By Lemma 2, turns out to be an approximation for , for some positive constant . Since is a sequence of elements from , there are such that . Then, , because is a homomorphism. Thus, is a sequence of elements from which approximates to an accuracy . For a given depth , is approximation error associated with the Solovay–Kitaev approximations in . Therefore, we can ensure that , which implies . Finally, this function returns a sequence of instructions from which approximates to an accuracy of .
4. Computational Experience
One challenge encountered in the computation is that fails to exist as the map is not one to one. This however was overcome using the fact that for given it is possible to find such that , for which the following construction was used. Any element in can be represented by a real number , the angle of rotation, and a rotation axis , which is a 3dimensional unit vector, denoted by . The corresponding matrix can be expressed explicitly by where
For a given rotational matrix , define
One can verify that and . Therefore, for an element , in order to find an element such that , under this construction, we need to find a unit vector and a real number such that .
Let , and suppose is the corresponding rotational matrix of (i.e., ). If is any vector parallel to , then it must satisfy , because the rotation of around the axis of rotation must result in . Since , we can always find an eigenvalue which is equal to 1, from which it immediately follows that is an eigenvector which corresponds to the eigenvalue 1. So by diagonalizing , we find the unit vector which is parallel to each other . Now, since both and are unit vectors, we must have or . By equation (18), the trace of the matrix reduces to , which immediately results in Now, by defining such that and choosing the right sign for to match the rotational axis (i.e., ), we get .
Accordingly, we implemented our algorithm in to find the Solovay–Kitaev approximates to several special unitary matrices. The computational experiment was conducted in accordance with the algorithmic steps mentioned and the bound was obeyed as in Theorem 5. We implemented with different instruction sets and the implementation with the instruction set in where
resulted in much shorter length for a given than the others. It would be an interesting future work to identify any classes or subgroups of matrices that can be approximated best by each instruction set, perhaps with a comparison of different instruction sets.
5. Discussion
Unconventional computing with different algebraic structures had been the topic of interest for a few previous works for which the primary motivation was quantum computing. Based on the fact that the circuitgate framework of quantum computing relies on the Solovay–Kitaev theorem, we investigated the possibility of deriving a version of this theorem for on a threelevel quantum system, indicating the potential of using orthogonal matrices for efficient computation.
Threelevel quantum systems and relevant operators had already been a topic of interest. In analogy with standard circuitgate framework, it was customary to use the elements in the unitary group as the operators in these systems. Despite the recent experimental achievements, theoretical bounds, capacity, and other related questions on threelevel quantum systems were seldom explored. With our version of the Solovay–Kitaev theorem, it is now known that efficient computation is possible with the orthogonal subgroup of . This is a noticeable distinction when compared with the subgroup of . Being an Abelian group, it is impossible to perform Solovay–Kitaev type approximations on . Thus, an instruction set in standard quantum computation enforces the inclusion of (the phase gate), , or , the nonorthogonal gates. However, the faulttolerant implementation of the phase gate is much more complicated than the orthogonal gates [25]. Therefore, quantum speedup only using orthogonals is beyond feasibility in standard circuitgate framework, though desired. In contrast to this, as our results indicate, quantum speedup with orthogonals is theoretically feasible in a threelevel quantum system.
It is worthwhile to consider our version of the Solovay–Kitaev theorem in the context quantum compilation [26, 27] in which the conversion of a nonfaulttolerant circuit into a fault tolerant one is investigated. A recent paper introduced several efficient methods for quantum compilation using physical machine descriptions, which included one method based on the Solovay–Kitaev approximations [28]. Although compilation and optimization of threelevel quantum circuits have been the subject of a few other works [19, 20, 29], none is based on the Solovay–Kitaev theorem. It would be an interesting future task to investigate whether efficient compilations are possible for a threelevel system with orthogonals. This would be possible by constructing a Solovay–Kitaevbased compilation method analogous to the one in [28], for which our algorithm in Section 3.1 would be helpful.
Our aim was particularly on exploring the approximation power of special orthogonal matrices. Therefore, we confined our study to a particular form of instruction sets in ; that is, the images of instruction sets in . A closer inspection reveals that an arbitrary instruction set in behaves similarly, resulting in the same length for a given accuracy. Therefore, it is an immediate consequence of Theorem 5 that slightly different versions of the Solovay–Kitaev theorem and algorithm for can be established. However, the applicability of the Solovay–Kitaev theorem to other Lie groups than still remains a nontrivial and theoretically interesting topic, which has not been investigated in literature. It would be a potential future task to see if the theorem is extendable to those groups.
Data Availability
No data were used to support this study.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
AM would like to thank Jingbo Wang, Lyle Noakes, André Nies, and Willem Fouché for insightful discussions.
References
 R. Solovay, Proof of Solovay Kitaev Theorem, 1995.
 A. Yur'evich Kitaev, “Quantum computations: algorithms and error correction,” Uspekhi Matematicheskikh Nauk, vol. 52, no. 6, pp. 53–112, 1997. View at: Google Scholar
 L. F. Willem, “An algorithmic construction of quantum circuits of high descriptive complexity,” Electronic Notes in Theoretical Computer Science, vol. 221, pp. 61–69, 2008. View at: Google Scholar
 A. N. Michael and I. Chuang, Quantum Computation and Quantum Information, 2000.
 D. Aharonov, W. Van Dam, J. Kempe, Z. Landau, S. Lloyd, and O. Regev, “Adiabatic quantum computation is equivalent to standard quantum computation,” SIAM Review, vol. 50, no. 4, pp. 755–787, 2008. View at: Publisher Site  Google Scholar
 A. M. Childs, D. Gosset, and Z. Webb, “Universal computation by multiparticle quantum walk,” Science, vol. 339, no. 6121, pp. 791–794, 2013. View at: Publisher Site  Google Scholar
 A. C.C. Yao, “Quantum circuit complexity,” pp. 352–361, IEEE. View at: Google Scholar
 D. Aerts and M. Czachor, “Cartoon computation: quantumlike computing without quantum mechanics,” Journal of Physics A: Mathematical and Theoretical, vol. 40, no. 13, pp. F259–F266, 2007. View at: Publisher Site  Google Scholar
 M. Czachor, “Elementary gates for cartoon computation,” Journal of Physics A: Mathematical and Theoretical, vol. 40, no. 31, pp. F753–F759, 2007. View at: Publisher Site  Google Scholar
 J. M. Fernandez and W. A. Schneeberger, “Quaternionic computing,” 2003, http://arxiv.org/abs/quantph/0307017. View at: Google Scholar
 M. A. Graydon, “Quaternionic quantum dynamics on complex hilbert spaces,” Foundations of Physics, vol. 43, no. 5, pp. 656–664, 2013. View at: Publisher Site  Google Scholar
 P. L. Ben, T. J. Weinhold, N. K. Langford et al., “Manipulating biphotonic qutrits,” Physical review letters, vol. 100, no. 6, article 060504, 2008. View at: Google Scholar
 P. Gokhale, J. M. Baker, C. Duckering, F. T. Chong, N. C. Brown, and K. R. Brown, “Extending the Frontier of Quantum Computers With Qutrits,” IEEE Micro, vol. 40, no. 3, pp. 64–72, 2020. View at: Publisher Site  Google Scholar
 Y.H. Luo, H.S. Zhong, M. Erhard et al., “Quantum teleportation in high dimensions,” Physical review letters, vol. 123, no. 7, p. 070505, 2019. View at: Publisher Site  Google Scholar
 B. Li, Z.H. Yu, and S.M. Fei, “Geometry of quantum computation with qutrits,” Scientific reports, vol. 3, no. 1, p. 2594, 2013. View at: Publisher Site  Google Scholar
 D. Mc Hugh and J. Twamley, “Trappedion qutrit spin molecule quantum computer,” New journal of physics, vol. 7, no. 1, p. 174, 2005. View at: Publisher Site  Google Scholar
 G. Khanna, S. Mukhopadhyay, R. Simon, and N. Mukunda, “Geometric phases for SU(3) representations and three level quantum systems,” Annals of Physics, vol. 253, no. 1, pp. 55–82, 1997. View at: Publisher Site  Google Scholar
 A. B. Klimov, R. Guzman, J. C. Retamal, and C. Saavedra, “Qutrit quantum computer with trapped ions,” Physical Review A, vol. 67, no. 6, article 062313, 2003. View at: Publisher Site  Google Scholar
 R. NaderAli, A. JafariDolama, and M. AmniatTalab, “Implementation of singlequtrit quantum gates via tripod adiabatic passage,” International Journal of Optics and Photonics, vol. 4, no. 1, pp. 39–48, 2010. View at: Google Scholar
 V. E. Zobov and V. P. Shauro, “On timeoptimal nmr control of states of qutrits represented by quadrupole nuclei with the spin i=1,” Journal of Experimental and Theoretical Physics, vol. 113, no. 2, pp. 181–191, 2011. View at: Publisher Site  Google Scholar
 S. Dogra, K. Dorai, and Arvind, “Majorana representation, qutrit hilbert space and nmr implementation of qutrit gates,” Journal of Physics B: Atomic, Molecular and Optical Physics, vol. 51, no. 4, p. 045505, 2018. View at: Publisher Site  Google Scholar
 J. F. Cornwell, Group Theory in Physics: An Introduction, Academic press, San Diego, California, United States., 1997.
 C. M. Dawson and M. A. Nielsen, “The Solovay Kitaev algorithm,” 2005, http://arxiv.org/abs/quantph/0505030. View at: Google Scholar
 E. Wigner, Group Theory and Its Application to the Quantum Mechanics of Atomic Spectra, 2012.
 P. Aliferis, D. Gottesman, and J. Preskill, “Quantum accuracy threshold for concatenated distance3 codes,” Quantum information & computation, vol. 6, pp. 97–165, 2005. View at: Google Scholar
 L. Biswal, D. Bhattacharjee, A. Chattopadhyay, and H. Rahaman, “Techniques for faulttolerant decomposition of a multicontrolled Toffoli gate,” Physical Review A, vol. 100, no. 6, 2019. View at: Publisher Site  Google Scholar
 A. Paler, I. Polian, K. Nemoto, and S. J. Devitt, “Faulttolerant, highlevel quantum circuits: form, compilation and description,” Quantum Science and Technology, vol. 2, no. 2, p. 025003, 2017. View at: Publisher Site  Google Scholar
 C.C. Lin, A. Chakrabarti, and N. K. Jha, “Ftqls: faulttolerant quantum logic synthesis,” IEEE Transactions on very large scale integration (VLSI) systems, vol. 22, no. 6, pp. 1350–1363, 2013. View at: Google Scholar
 T. Bækkegaard, L. B. Kristensen, N. J. S. Loft, C. K. Andersen, D. Petrosyan, and N. T. Zinner, “Realization of efficient quantum gates with a superconducting qubitqutrit circuit,” Scientific Reports, vol. 9, no. 1, p. 13389, 2019. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2020 Anuradha Mahasinghe et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.