---
title: Continuous-Time Nonlinear Quantum Walk
url: https://www.emergentmind.com/topics/continuous-time-nonlinear-quantum-walk
type: topic
---

# Continuous-Time Nonlinear Quantum Walk

A continuous-time nonlinear quantum walk is a quantum walk whose evolution is governed by a nonlinear Schrödinger equation rather than a linear one. In the one-dimensional setting studied for paths and cycles, the effective Hamiltonian combines nearest-neighbor hopping with an on-site cubic term of Gross–Pitaevskii type,
\[
H \;=\; -J\,\sum_{\langle n,m\rangle}\Bigl(\lvert n\rangle\langle m\rvert+\lvert m\rangle\langle n\rvert\Bigr)
\;+\; g\sum_{n=0}^{N-1}|\psi_n|^2\,\lvert n\rangle\langle n\rvert,
\]
where \(J>0\) is the hopping rate and \(g\in\mathbb R\) is the nonlinearity coefficient [2605.20464]. The resulting dynamics differs qualitatively from linear one-dimensional quantum walks: instead of rapid spreading, the walker can remain localized at the initially occupied site to arbitrary fidelity depending on the nonlinear coefficient, a phenomenon proved analytically for discrete paths and cycles with cubic nonlinearity [2605.20464]. Closely related continuous-time models also underlie nonlinear spatial search on incomplete graphs, where on-site nonlinear self-potentials can accelerate search relative to the linear quantum-walk regime [2606.01403].

## 1. Formal definition and governing equations

For a walker on a one-dimensional graph of \(N\) sites \(v_0,\dots,v_{N-1}\), with site basis \(\{\lvert n\rangle:n=0,\dots,N-1\}\), the state is written as
\[
\lvert\Psi(t)\rangle \;=\;\sum_{n=0}^{N-1}\psi_n(t)\,\lvert n\rangle,
\qquad
\sum_n|\psi_n|^2=1.
\]
With \(\hbar=1\) and often \(J=1\), the Schrödinger equation becomes the discrete Gross–Pitaevskii equation
\[
i\,\frac{d\psi_n}{dt}
\;=\;
-\sum_{m:\,m\sim n}\psi_m
\;-\;
g\,|\psi_n|^2\,\psi_n,
\qquad
n=0,\dots,N-1,
\]
where \(m\sim n\) denotes neighbors of \(n\) [2605.20464].

The cubic term acts as an on-site potential proportional to the instantaneous probability. In the stated physical interpretation, such nonlinearities arise in Bose–Einstein condensates described by the Gross–Pitaevskii equation or by nonlinear optical waveguide arrays [2605.20464]. In this formulation, the nonlinearity is local, state-dependent, and diagonal in the site basis.

A broader continuous-time nonlinear quantum-walk framework appears in spatial search. There the evolution is again nonlinear,
\[
i\frac{d}{dt}\ket{\psi(t)}
\;=\;
H(t)\,\ket{\psi(t)},
\qquad
H(t)=H_0-V_{\rm nl}(t),
\]
with
\[
H_0=-\gamma L-\ket{w}\bra{w},
\qquad
V_{\rm nl}(t)
=
g \sum_{n=1}^N
f\!\bigl(\bigl|\braket{n}{\psi(t)}\bigr|^2\bigr)\,
\ket{n}\bra{n},
\]
where \(L=A-D\) is the graph Laplacian, \(\gamma\) is a hopping rate, \(\ket{w}\bra{w}\) is the oracle marking the target vertex, and \(f(p)\) specifies the nonlinearity [2606.01403]. Two forms are emphasized: the cubic nonlinearity \(f(p)=p\), which reproduces the Gross–Pitaevskii-type self-potential, and the cubic–quintic nonlinearity \(f(p)=p-p^2\) [2606.01403].

## 2. Self-trapping on one-dimensional paths and cycles

The central analytical result for one-dimensional continuous-time nonlinear quantum walks is self-trapping. For a walk started at a single vertex \(r\), let \(a(t)=\psi_r(t)\) and \(p(t)=|a(t)|^2\), with the remaining amplitudes collected into a vector \(b(t)\) satisfying \(\|b\|^2=1-p\). The proof of trapping proceeds through a conserved functional,
\[
E[\Psi]
=\langle\Psi|\,H\,-\,\tfrac12\,g\,|\Psi|^2\,|\Psi\rangle
=
\langle\Psi|H|\Psi\rangle
-
\tfrac g2\sum_n|\psi_n|^4,
\]
which is constant in time. At \(t=0\), when all amplitude is at \(r\), one has \(E=-g/2\), so
\[
\langle\Psi|H|\Psi\rangle
=
-\frac g2
+\frac g2\sum_n|\psi_n|^4
\]
for all \(t\) [2605.20464].

A second ingredient is a bound on the energy expectation,
\[
\bigl|\langle\Psi|H|\Psi\rangle\bigr|
\;\le\;
2\,\sqrt{p(1-p)\,\mathrm{deg}(v_r)}
\;+\;2\,(1-p),
\]
obtained by viewing the adjacency-matrix block coupling \(a\) and \(b\), then applying Cauchy–Schwarz together with the spectral norm of the remaining block [2605.20464]. Combining these relations yields
\[
|g|\,p\,(1-p)\;\le\;
2\,\sqrt{p(1-p)\,\mathrm{deg}(v_r)}
\;+\;2\,(1-p),
\]
which is rearranged as
\[
f(p)\;=\;
\frac{2\sqrt{\mathrm{deg}(v_r)}\,\sqrt{p(1-p)}\;+\;\tfrac{2}{\,p\,}}
{}\;\ge\;|g|.
\]
The function \(f(p)\) diverges as \(p\to0\) and \(p\to1\) and has a unique minimum \(f_{\mathrm{min}}\in(0,1)\). If \(|g|>f_{\mathrm{min}}\), the trajectory \(p(t)\) cannot cross the two roots \(p_-<p_+\) of \(f(p)=|g|\). Since \(p(0)=1\), one obtains
\[
\forall\,t\ge0:\quad p(t)\ge p_+,
\]
so the probability at the initially occupied site is bounded below by \(p_+\), establishing self-trapping [2605.20464].

For the threshold values determined by degree alone, the rigorous minima are \(f_{\mathrm{min}}\approx7.22\) when \(r\) is an endpoint of a path, so \(\deg(r)=1\), and \(f_{\mathrm{min}}=9\) when \(r\) is an interior site of a path or any site on a cycle, so \(\deg(r)=2\) [2605.20464]. The same source states that numerically the true trapping threshold is \(|g_c|\approx4\), while the proof yields non-tight bounds \(|g_c|\le7.22\) or \(\le9\) [2605.20464]. As \(g\to\infty\), \(p_+\to1\), and the walker becomes arbitrarily well-trapped; for moderate \(|g|\), \(p_+\approx1-O(1/|g|)\) [2605.20464].

This establishes a precise contrast with linear one-dimensional quantum walks, which are known for spreading quickly in one dimension [2605.20464]. A plausible implication is that the nonlinear term effectively converts one-dimensional transport from a ballistic-spreading regime into a tunably localized regime.

## 3. Dependence on graph structure and initial placement

On the finite path \(P_N\), the dynamics depends on whether the walker starts at an endpoint or in the interior. If the walker starts at an endpoint, where \(\deg=1\), trapping appears for \(|g|\gtrsim4\) numerically, while the proof requires \(g>7.22\). If the walker starts in the interior, where \(\deg=2\), trapping again appears for \(|g|\gtrsim4\) numerically, while the proof requires \(g>9\) [2605.20464].

On the cycle \(C_N\), all sites have \(\deg=2\), so the same interior threshold applies [2605.20464]. In the analytical framework of the proof, the degree of the initially occupied vertex is the sole graph-theoretic parameter entering the lower-bound criterion through \(\mathrm{deg}(v_r)\). This suggests that, within this one-dimensional class, the local coordination number controls the rigorous trapping bound more directly than the global graph size.

The spatial-search setting broadens the graph dependence in a different direction. There, nonlinear speedups are established for graphs that are “sufficiently complete,” meaning that under linear search the system effectively evolves in a two-dimensional subspace spanned by \(\ket{s}\pm\ket{w}\), with asymptotic spectral gap \(\Delta E=2/\sqrt{N}[1+O(\epsilon)]\) for some \(\epsilon\to0\) [2606.01403]. The examples explicitly listed are Paley graphs and complete bipartite graphs with \(N_1,N_2=\Theta(N)\), both having three symmetry classes of vertices \(m_0,m_1,m_2\) and \(\epsilon=1/\sqrt{N}\) [2606.01403]. Hypercubes and arbitrary-dimensional cubic lattices are treated numerically rather than through the same closed-form proof [2606.01403].

The two one-dimensional trapping problem and the spatial-search problem therefore emphasize different structural features. In the former, the decisive parameter in the proof is the local degree of the starting vertex; in the latter, the decisive condition is the approximate two-dimensional spectral structure of the graph under linear search [2605.20464; 2606.01403].

## 4. Quantum state transfer and quantum memory interpretation

The one-dimensional self-trapping effect is proposed as a timing mechanism for quantum state transfer. The stated protocol on a small spin chain such as \(P_3\) is: initialize the walker, encoding the qubit, at the source node \(r\); turn on a large nonlinearity \(|g|\gg f_{\mathrm{min}}\) to trap the amplitude at \(r\) with fidelity at least \(p_+(g)\); when ready to transfer, suddenly set \(g=0\); allow linear evolution to perform perfect or pretty-good state transfer to the target node in a known time \(T_{\rm PST}\); and upon arrival, restore \(|g|\) large to trap the qubit at the receiving node indefinitely until use [2605.20464].

The specific example given is the path \(P_3\), for which perfect state transfer from \(0\to2\) occurs at
\[
T=\pi/\sqrt2.
\]
The same summary states that one can hold with \(g=40\) for arbitrary waiting times before and after the linear transfer window [2605.20464].

Within this scheme, the same chain is interpreted as both storage and transmission line. The storage fidelity is the self-trapping bound \(p_+(g)\), and the release fidelity is the fidelity of the underlying linear state transfer [2605.20464]. The source explicitly describes this as a “quantum memory viewpoint,” with trapping and transfer corresponding respectively to storage and release of quantum information [2605.20464].

The proposed candidate implementations are Bose–Einstein condensates in optical lattices, where the Gross–Pitaevskii nonlinearity \(g\) is proportional to the scattering length and tunable via Feshbach resonance, and nonlinear optical waveguide arrays with Kerr nonlinearity, where \(g\) can be modulated by pump intensity or external fields [2605.20464]. A plausible implication is that the timing protocol depends not only on the existence of a nonlinear trapping regime but also on the ability to switch or modulate the effective nonlinearity on operational timescales.

## 5. Nonlinear spatial search and runtime scaling

In continuous-time spatial search, the walker begins in the uniform superposition
\[
\ket{\psi(0)}=\ket{s}
=\frac{1}{\sqrt{N}}
\sum_{n=1}^N\ket{n},
\]
and the marked vertex \(w\) enters through the oracle term \(-\ket{w}\bra{w}\) in the linear part of the Hamiltonian [2606.01403]. In the linear case, one chooses a critical hopping rate \(\gamma_L\) so that two eigenvalues of \(H_0\) become nearly degenerate with gap \(\Delta E\sim2/\sqrt N\). In the nonlinear regime, this is generalized to a time-dependent rate
\[
\gamma_c(t)
=
\gamma_L\,
\Bigl[1
+g\bigl(f_0(t)-f_\ell(t)\bigr)\Bigr],
\]
where \(f_k(t)=f(|c_k(t)|^2/|m_k|)\) are the nonlinear self-potentials on symmetry classes \(m_k\); this choice forces the nonlinear evolution to track the linear path in a rescaled time [2606.01403].

For cubic nonlinearity \(f(p)=p\), the theorem stated for sufficiently complete graphs assumes
\[
g\epsilon=o(1),\qquad
gB_\ell\sqrt N=o(1),\qquad
g^2\epsilon B_\ell\sqrt N=o(1),
\]
with \(B_\ell=\max_{k\ne0,\ell}(1/|m_k|+1/|m_\ell|)\). Under these conditions, up to the first success time the nonlinear evolution remains within \(o(1)\) of the rescaled linear critical-search path, and the hitting time satisfies
\[
t_*
=
\frac{\pi\sqrt N}{2\sqrt{1+g}\,[1+o(1)]}.
\]
For fixed \(g>0\), this remains \(O(\sqrt N)\), while for \(g=\Theta(N)\) it becomes \(O(1)\), giving constant-time search [2606.01403].

Applied to Paley graphs or complete bipartite graphs with \(N_1,N_2=\Theta(N)\), the same summary states that for any \(g\ll\sqrt N\) the conditions hold and
\[
t_*=\frac{\pi\sqrt N}{2\sqrt{1+g}[1+o(1)]}.
\]
A refined analysis for Paley graphs permits \(g=N-1\), in which case the nonlinear algorithm succeeds in constant time \(t_*=\Theta(1)\) with probability \(1-O(1/N)\) [2606.01403].

For the cubic–quintic nonlinearity \(f(p)=p-p^2\), the corresponding runtime is
\[
t_*=\frac{\pi\sqrt N}{2\sqrt{1+g/4}\,[1+o(1)]},
\]
so \(g=\Theta(N)\) again yields constant-time search [2606.01403]. Numerical evidence extends this behavior to hypercubes of dimension \(n\), where simulations with \(g=N-1\) show nearly constant \(t_*\approx O(1)\) and success probability near unity for \(N=2^7,2^{10}\), and to periodic \(d\)-dimensional lattices, where cubic or cubic–quintic nonlinearity with \(g=N-1\) gives near-constant \(t_*\) and high success for \(d\ge5\), but fails to reach the same success probability as the linear case for \(d=2,3,4\) [2606.01403]. The threshold dimension is therefore reported as \(d_{\rm crit}=5\) [2606.01403].

## 6. Relation to nonlinear Dirac limits, photonic models, and open issues

Although the one-dimensional self-trapping and spatial-search models are continuous-time systems from the outset, related nonlinear quantum-walk literature often studies discrete-time walks whose continuum limit is a nonlinear Dirac equation. In a general nonlinear quantum walk on the lattice \(\delta\mathbb Z\), the one-step evolution operator is
\[
U_\delta=S_\delta C_\delta N_\delta,
\]
with a shift \(S_\delta\), a linear coin \(C_\delta\), and a nonlinear state-dependent coin \(N_\delta\). Using Shannon interpolation, the walker on \(\delta\mathbb Z\) is shown to converge uniformly in Sobolev space \(H^s\) on fixed time intervals to the solution of a nonlinear Dirac equation as \(\delta\to0\) [1902.02017]. The limiting equation is
\[
i\,\partial_t\,u
=
-\;i\,\sigma_3\,\partial_x\,u
+
s(x)\cdot\sigma\,u
+
g(\langle u,\gamma u\rangle)\,\gamma\,u,
\]
and the convergence theorem states
\[
\sup_{0\le m\le T/\delta}
\bigl\|
I_\delta U_\delta^m I_\delta^{-1}j_\delta u_0
-
U_{\rm Dirac}(m\delta)u_0
\bigr\|_{H^s}
\le C\delta
\]
under the stated regularity and Lipschitz assumptions [1902.02017].

A different discrete nonlinear model, designed for photonic implementation using an optical nonlinear Kerr medium, yields in a space-time continuum limit a nonlinear Dirac equation with a nonlinear mass term,
\[
[i\gamma^\mu\partial_\mu-m(\Psi)]\Psi=0,
\qquad
m(\Psi)=\tilde\theta_0-\frac{\tilde\alpha}{2}\Psi^\dagger\sigma_y\Psi,
\]
and supports approximate bright and dark solitons in the continuum analysis [2308.01014]. Numerical iteration of the underlying map produces localized bright solitons, kick-activated motion, soliton–soliton collisions in which packets pass through one another largely unscathed, dark solitons, and modified behavior under synthetic electric fields [2308.01014].

These discrete-time continuum-limit results do not define the same model as the continuous-time Gross–Pitaevskii walk on a path or cycle. They nevertheless place continuous-time nonlinear quantum walks within a larger research program linking state-dependent walk dynamics, effective nonlinear field equations, and localized structures such as self-trapped states and solitons [1902.02017; 2308.01014]. A common misconception is that “nonlinear quantum walk” denotes a single canonical construction; the cited literature instead includes at least continuous-time Gross–Pitaevskii-type walks on graphs, continuous-time search Hamiltonians with nonlinear self-potentials, and discrete-time state-dependent coin walks with nonlinear Dirac limits [2605.20464; 2606.01403; 1902.02017; 2308.01014].

Several limitations and open questions are explicitly identified in the spatial-search setting. The effective nonlinear Schrödinger equation arises in the mean-field Gross–Pitaevskii limit of a Bose–Einstein condensate with two-body interactions, and the coupling \(g\) scales inversely with the number of particles \(N_{\rm part}\), so achieving large \(g\) while maintaining the mean-field regime requires a trade-off between condensate size and interaction strength [2606.01403]. The same source notes that low-dimensional lattices and sparse graphs that fail to approximate the two-dimensional spectral structure of the complete graph do not benefit except at very large \(g\) that break the approximations, and it lists open questions concerning rigorous hypercube speedup proofs, universal lower bounds on \(g\) versus graph spectral properties, and particle-number versus connectivity trade-offs in a Bose–Einstein-condensate implementation [2606.01403].

Source: https://www.emergentmind.com/topics/continuous-time-nonlinear-quantum-walk