Abstract
In this paper, we present closed-form expressions for the wave function that governs the evolution of the discrete-time quantum walk on the line when the coin operator is arbitrary. The formulas were derived assuming that the walker can either remain put in the place or proceed in a fixed direction but never move backward, although they can be easily modified to describe the case in which the particle can travel in both directions. We use these expressions to explore properties of magnitudes associated to the process, as the probability mass function or the probability current, even though we also consider the asymptotic behavior of the exact solution. Within this approximation, we will estimate upper and lower bounds, examine the origins of an emerging approximate symmetry, and deduce the general form of the stationary probability density of the relative location of the walker.
Similar content being viewed by others
Notes
Some of the intermediate expressions shown along this paper may be ill-defined when either \(\theta =0\) or \(\theta =\frac{\pi }{2}\). In spite of that, the final formulas we obtain are still valid in the full range.
In Ref. [30] we discussed the virtues of the alternative choice \(N=2^m\), with \(m\) the smallest integer for which it holds \(t<2^m\). With this setting one can resort to the fast Fourier transform to perform all the calculations, thereby greatly reducing the overall computational time, a point that may be very relevant for large values of \(t\).
We show below how \(|\psi _{1}(1,1)|^2\) can be understood as the (rightward) “initial velocity” of our walker. In a bidirectional scheme, \(|\psi _{0}(0,1)|^2\) would play the role of the leftward “initial velocity.”
This idea of expressing the evolution of the walker in terms of a unique magnitude which follows a recursive equation is not new, e.g., in Ref. [3], we find one of such recursion formulas which involves non-constant coefficients. Especially relevant is, in this sense, Ref. [23] where it is derived the bidirectional equivalent of Eq. (31) on the basis of the properties of the Chebyshev polynomials of the second kind.
For that reason \(\bar{\rho }_\mathrm{med}(n,t)\) and \(\bar{\rho }_\mathrm{osc}(n,t)\) are sometimes referred as the slow component and the fast component of \(\bar{\rho }(n,t)\), respectively [2].
For notational convenience, \(N\) and \(T\) may alternate or even coexist in expressions in this appendix.
References
Aharonov, Y., Davidovich, L., Zagury, N.: Quantum random walks. Phys. Rev. A 48, 1687–1690 (1993)
Nayak, N., Vishwanath, A.: Quantum Walk on the Line. arXiv:quant-ph/0010117 (2000)
Ambainis, A., Bach, E., Nayak, A., Vishwanath, A., Watrous, J.: One dimensional quantum walks. In: Proceedings of the Thirty-Third Annual ACM Symposium on Theory of Computing. ACM New York, New York, pp. 37–49 (2001)
Travaglione, B.C., Milburn, G.J.: Implementing the quantum random walk. Phys. Rev. A 65, 032310 (2002)
Konno, N.: Quantum random walks in one dimension. Quantum Inf. Process. 1, 345–354 (2003)
Kempe, J.: Quantum random walks: an introductory overview. Contemp. Phys. 44, 307–327 (2003)
Venegas-Andraca, S.E.: Quantum walks: a comprehensive review. Quantum Inf. Process. 11, 1015–1106 (2012)
Shor, P.W.: Polynomial-time algorithms for prime factorization and discrete logarithms on a quantum computer. SIAM J. Comput. 26, 1484–1509 (1997)
Farhi, E., Gutmann, S.: Quantum computation and decision trees. Phys. Rev. A 58, 915–928 (1998)
Childs, A., Farhi, E., Gutmann, S.: An example of the difference between quantum and classical random walks. Quantum Inf. Process. 1, 35–43 (2003)
Shenvi, N., Kempe, J., Whaley, K.B.: Quantum random-walk search algorithm. Phys. Rev. A 67, 052307 (2003)
Agliari, E., Blumen, A., Nülken, O.: Quantum-walk approach to searching on fractal structures. Phys. Rev. A 82, 012305 (2010)
Magniez, F., Nayak, A., Roland, J., Santha, M.: Search via quantum walk. SIAM J. Comput. 40, 142–164 (2011)
Flitney, A.P., Abbott, D., Johnson, N.F.: Quantum walks with history dependence. J. Phys. A 37, 7581–7591 (2004)
Bulger, D., Freckleton, J., Twamley, J.: Position-dependent and cooperative quantum Parrondo walks. New J. Phys. 10, 093014 (2008)
Chandrashekar, C.M., Banerjee, S.: Parrondo’s game using a discrete-time quantum walk. Phys. Lett. A 375, 1553–1558 (2011)
Tregenna, B., Flanagan, W., Maile, R., Kendon, V.: Controlling discrete quantum walks: coins and initial states. New J. Phys. 5, 83 (2003)
Bach, E., Coppersmith, S., Goldschen, M.P., Joynt, R., Watrous, J.: One-dimensional quantum walks with absorbing boundaries. J. Comput. Syst. Sci. 69, 562–592 (2004)
Chandrashekar, C.M., Srikanth, R., Laflamme, R.: Optimizing the discrete time quantum walk using a SU(2) coin. Phys. Rev. A 77, 032326 (2008)
Chandrashekar, C.M., Srikanth, R., Banerjee, S.: Symmetries and noise in quantum walk. Phys. Rev. A 76, 022316 (2007)
Asbóth, J.K.: Symmetries, topological phases, and bound states in the one-dimensional quantum walk. Phys. Rev. B 86, 195414 (2012)
Kitagawa, T.: Topological phenomena in quantum walks: elementary introduction to the physics of topological phases. Quantum Inf. Process. 11, 1107–1148 (2012)
Fuss, I., White, L., Sherman, P., Naguleswaran, S.: An analytic solution for one-dimensional quantum walks. arXiv:0705.0077 (2007)
Villagra, M., Nakanishi, M., Yamashita, S., Nakashima, Y.: Quantum walks on the line with phase parameters. IEICE Trans. Inf. Syst. E95.D, 722–730 (2012)
Grimmett, G., Janson, S., Scudo, P.F.: Weak limits for quantum random walks. Phys. Rev. E 69, 026119 (2004)
Konno, N.: A new type of limit theorems for the one-dimensional quantum random walk. J. Math. Soc. Jpn. 57, 1179–1195 (2005)
Bressler, A., Pemantle, R.: Quantum random walks in one dimension via generating functions. In: Proceedings of the 2007 Conference on Analysis of Algorithms, pp. 403–414 (2007)
Ahlbrecht, A., Vogts, H., Werner, A.H., Werner, R.F.: Asymptotic evolution of quantum walks with random coin. J. Math. Phys. 52, 042201 (2011)
Hoyer, S., Meyer, D.A.: Faster transport with a directed quantum walk. Phys. Rev. A 79, 024307 (2009)
Montero, M.: Unidirectional quantum walks: evolution and exit times. Phys. Rev. A 88, 012333 (2013)
Hillery, M., Bergou, J., Feldman, E.: Quantum walks based on an interferometric analogy. Phys. Rev. A 68, 032314 (2003)
Courant, R., Hilbert, D.: Methods of Mathematical Physics. Cambridge University Press, Cambridge (1953)
Romanelli, A.: Thermodynamic behavior of the quantum walk. Phys. Rev. A 85, 012319 (2012)
Romanelli, A., Segundo, G.: The entanglement temperature of the generalized quantum walk. Phys. A 393, 646–654 (2014)
Acknowledgments
The author acknowledges partial support from the former Spanish Ministerio de Ciencia e Innovación under Contract No. FIS2009-09689, from the Spanish Ministerio de Economía y Competitividad under Contract No. FIS2013-47532-C3-2-P, and from Generalitat de Catalunya, Contract No. 2014SGR608.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1: Exact solution
In this appendix, we provide explicit details on the derivation of the closed-form expressions for the two components of the wave function given in the main text, Eqs. (20) and (21), starting from the recursive formulas (18) and (19). The approach that follows is similar to the one taken in our previous work [30] and differs from the broadest method in the use of the discrete Fourier transform (DFT) instead of the discrete-time Fourier transform [7].
Let \(f(n)\) be a given set of \(N\) complex numbers, \(n~\in ~\{0,\ldots , N-1\}\), and denote its DFT by \(\tilde{f}(r)\),
for \(r\in \{0,\ldots , N-1\}\). The \(N\) complex quantities \(\tilde{f}(r)\) thus defined keep exactly the same information contained in the original series, and therefore, one can recover \(f(n)\) from \(\tilde{f}(r)\) by means of the so-called inverse DFT formula
The DFT is well suited to convert recursive relationships in the position domain, as Eqs. (18) and (19), into a set of algebraic equations in the Fourier domain. In our case, however, the recursive formulas involve not only different locations but different instants of time, so we must carefully choose \(f(n)\) and \(N\) itself to preserve the overall coherence.
To this end, let us introduce the auxiliary time horizon \(T\), \(T\ge 0\), set \(N\equiv T+1\), and consider the following definition for the DFT of \(\psi _{0,1}(n,t)\), valid for any \(t\), \(t~\in ~\{0,\ldots , T\}\), Footnote 6
where it is implicitly assumed that \(\psi _{0,1}(n,t)=0\) for any \(n\ge t+1\). This definition entails that \(\tilde{\psi }_{0,1}(r,t;T)\) is an explicit function of \(T\): that is, for a fixed value of \(r\) and a fixed value of \(t\), different choices of \(T\) lead to different values for \(\tilde{\psi }_{0,1}(r,t;T)\). Nonetheless, the final result that one gets after applying the corresponding inversion formula,
does not depend on \(T\) for a fixed choice of \(n\) and \(t\), as long as \(0\le n\le t\le N-1\). Footnote 7 Therefore, the conclusion is that Eq. (71) will be valid for any \(N\), \(N\ge t+1\).
At this point, we can safely move Eqs. (18) and (19) into the Fourier domain:
The initial values for \(\tilde{\psi }_{0,1}(r,t;T)\) are \(\tilde{\psi }_{0}(r,0;T)=\cos \eta \), \(\tilde{\psi }_{1}(r,0;T)=\sin \eta \), for \(r \in \{0,\ldots ,N-1\}\). The resolution of Eqs. (72) and (73) can be tackled through standard matrix techniques, thus resulting in
and
with \(\lambda _{\pm }\) functions of \(r/N\),
and where \(\omega _{\frac{r}{N}}\) is an angle that, given \(r\) and \(N\), satisfies
Note that, since \(r\in \{0,\ldots ,N-1\}\), we have
so, to prevent any uncertainty, we consider that \(\omega _{\frac{r}{N}}\) is the only solution that Eq. (78) has in \([0,\frac{\pi }{2}]\).
Now, we can simply introduce the value of \(\tilde{\psi }_{0,1}(r,t;T)\) given in (74) and (75) into Eq. (71) and recover \(\psi _{0,1}(n,t)\) after the computation of a finite sum. Note that, at this point, our procedure has yielded a closed-form expression for the wave function; thus, we could simply stop here. However, we are going to simplify the final formulas as much as possible: In this way, we can get a more detailed view of the properties of the solution.
To manage the complexity of this endeavor we analyze the particular case \(\eta =0\) first. Let us begin with \(\psi _{0}(n,t)\):
If we use Eq. (78) in conjunction with the definition of \(\lambda _{\pm }\), Eqs. (76) and (77), we can show that
holds, and that consequently
Here, we have isolated in the first term all the contribution coming from the \(r=0\) case. This step is necessary to rearrange the two last summations by introducing a new variable \(s\), \(s\equiv N-r\),
where the following identities
have been taken into account. Finally, we find
In the case of \(\psi _{1}(n,t)\), we can proceed in a similar way. The result
is almost immediate once one realizes that
an expression that combines Eqs. (78) and (80).
Analogously, when \(\eta =\frac{\pi }{2}\) we obtain
and
by simply applying the same ideas and intermediate formulas.
Finally, we can recover the general solution, Eqs. (20) and (21), through the superposition of these two cases.
Appendix 2: Asymptotic expressions
In this appendix, we obtain asymptotic expressions for \(\psi _{0,1}(n,t)\) and \(\rho (n,t)\), formulas with a restricted validity but which in turn are more compact and readable than the exact ones.
We begin with a close analysis of the inner structure of the different pieces that compose Eqs. (20) and (21). The conclusion is that, in essence, we must find a way to approximate functions like \(h(n,t)\),
where
and \(\upsilon \in \{-1,0,1\}\). In every case, \(g(\cdot )\) is a smooth function, and therefore, the behavior of the cosine terms does determine the overall result of the sum. Due to the presence of \(\omega _{\frac{r}{N}}\) within \(\varPhi (n,r,t;T)\), the argument of these cosine functions does not change linearly with \(r\) but exhibits a maximum, and then, the use an adapted version of the method of the stationary phase is the one most indicated in this case [30, 32]: Only those terms for which \(\varPhi (n,r,t;T)\) attains its maximum are relevant, whereas the rest of them are negligible.
To this end, let us first define \(u\equiv r/N\), and \(\nu \equiv n/t\), in terms of which we can rewrite \(\varPhi (n,r,t;T)\),
with
Our next step is to consider the function \(h(n,t)\) in the continuum limit, \(N\rightarrow \infty \),
and to expand \(\phi (\nu ,u)\) in the vicinity of \(u_0\),
where \(u_0\) is the point for which, given \(\nu \), \(\phi (\nu ,u)\) has its maximum:
From Eq. (92), we have
and
Equation (94) shows us that the validity of the present approximation is restricted to values of \(\nu \) for which one has \(\cos ^2\theta -(2\nu -1)^2\ge 0\); that is,
since, by construction, \(\cos \theta \ge 0\). Also from Eqs. (22) and (94), we get
expressions that will be helpful in forthcoming derivations.
Now, we can fully evaluate Eq. (91) under the above premises:
with
The approximate versions of Eqs. (20) and (21) are
and
and they follow from Eq. (97) once one realizes, cf. Eqs. (93) and (96), that if
one gets
if
one has
and finally, if
one obtains
To derive the asymptotic expression for \(\rho (n,t)\), one has to calculate \(\left| \psi _{0,1}(n,t)\right| ^2\) first:
and
then use the following trigonometric identities,
in coordination with Eqs. (93) to (96), to ultimately obtain
This expression leads to Eq. (48) after a straightforward trigonometric transformation.
Note finally that, beyond being cornerstones in the calculation of \(\bar{\rho }(n,t)\), Eqs. (99) and (100) are worth by themselves: There are problems that require precise information about the asymptotic behavior of each wave function, as in [33, 34].
Rights and permissions
About this article
Cite this article
Montero, M. Quantum walk with a general coin: exact solution and asymptotic properties. Quantum Inf Process 14, 839–866 (2015). https://doi.org/10.1007/s11128-014-0908-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11128-014-0908-6