Pantograph Difference Equations: Stability, Multistability, and Chaos

Prashant M. Gade and Sachin Bhalekar

arXiv: 2610.21635v1 · PDF · 15 October 2026
Categories: math.DS, nlin.CD
Licence: CC BY 4.0 The site shows the text.

Abstract

Pantograph equations are functional differential equations with proportional delay, typically involving terms of the form x(at), where 0<a<1. These equations have been studied for several decades and have applications in areas ranging from number theory to electrodynamics. In this work, we investigate pantograph difference equations as discrete dynamical maps of the form

x_{i+1}=F(x_{i})+G(x_{h(i)}), \qquad i\geq 0,

where h(i)\leq i. In particular, we consider h(i)=\lfloor ai \rfloor and h(i)=\lfloor ai^{\alpha}\rfloor, with 0<a<1 and 0<\alpha<1. We distinguish between homogeneous pantograph difference equations, for which F(x)=0, and inhomogeneous pantograph difference equations, for which F(x)\ne 0. We first analyze linear pantograph difference equations and establish sufficient conditions for asymptotic stability and for nonconvergence to zero. For proportional delays, analytical stability and instability regions are combined with multi-horizon parameter scans to construct regime diagrams in the (\lambda,\delta)-plane. We then investigate pantograph versions of the logistic and tent maps. In the homogeneous case, the dynamics evolves through step-like plateaus whose lengths increase at least exponentially, leading to slow convergence and finite-time multistability within computationally accessible time scales. In the inhomogeneous case, numerical experiments show finite-time exponential separation of nearby trajectories, consistent with chaotic dynamics.

Items 11

Theorem

Theorem 1

6JZ1

Suppose that h(n)\to\infty as n\to\infty. The zero solution of the linear delay difference equation (?) is asymptotically stable if |\lambda| + |\delta| < 1.

Proof of Theorem 1

Set q=|\lambda|+|\delta|<1. Induction gives |x_n|\leq |x_0| for every n, so the solution is bounded and the zero solution is stable. Let

L=\limsup_{n\to\infty}|x_n|.

Because h(n)\to\infty, we also have \limsup_{n\to\infty}|x_{h(n)}|\leq L. Taking the upper limit in

|x_{n+1}|\leq |\lambda|\,|x_n|+|\delta|\,|x_{h(n)}|

therefore yields L\leq qL. Since q<1, it follows that L=0, and hence x_n\to0.

Note

14DM

The assumption h(n)\to\infty is essential. For example, if h(n)=0 and \delta\ne0, the persistent term \delta x_0 generally prevents convergence to zero. The condition |\lambda|+|\delta|<1 is sufficient but not necessary for proportional delays. Figure ? shows a numerically decaying orbit of (?) with \lambda=-0.50, \delta=0.52, and h(n)=\lfloor 0.2n\rfloor, even though |\lambda|+|\delta|>1.

Theorem

Theorem 2

6FQM

If \lambda,\delta\geq0 and \lambda+\delta>1, then the zero solution of (?) is not asymptotically stable for any delay function satisfying h(n)\leq n.

Used by Theorem 3

Proof of Theorem 2

Choose x_0>0. If x_j\geq x_0 for 0\leq j\leq n, then

x_{n+1}=\lambda x_n+\delta x_{h(n)}\geq(\lambda+\delta)x_0>x_0.

The assertion follows by induction: the solution cannot converge to zero.

Theorem

Theorem 3

CQ82

Under the hypotheses of Theorem 2, if \lambda>0, \delta>0, and h is monotone non-decreasing, then the solution with x_0>0 is strictly increasing.

Uses Theorem 2

Proof of Theorem 3

Assume x_0>0. Since \lambda>0 and \delta>0, all subsequent iterates are positive.

We show by induction that \{x_n\} is strictly increasing and hence does not converge to zero.

Indeed, since h(0)=0,

x_1=\lambda x_0 + \delta x_{h(0)}= \lambda x_0 + \delta x_0 > x_0.

Assume that x_n>x_{n-1} for n=1,2,\cdots,k. Since h is monotone non-decreasing, h(k)\geq h(k-1). Therefore,

x_{k+1}=\lambda x_k + \delta x_{h(k)} > \lambda x_{k-1} + \delta x_{h(k-1)}=x_k.

Thus, x_n is strictly increasing by induction.

Theorem

Theorem 4

C182

|\lambda|>1+|\delta|,

then the zero solution of (?) is unstable for every delay function satisfying h(n)\leq n. More precisely, every solution with x_0\ne0 is unbounded.

Proof of Theorem 4

Set q=|\lambda|-|\delta|>1. We prove inductively that

|x_{n+1}|\geq q|x_n|>|x_n|.

Since h(0)=0, the reverse triangle inequality gives

|x_1|=|\lambda+\delta|\,|x_0| \geq (|\lambda|-|\delta|)|x_0|=q|x_0|.

Now suppose that |x_0|<|x_1|<\cdots<|x_n|. Because h(n)\leq n, we have |x_{h(n)}|\leq |x_n|. Hence

\begin{aligned} |x_{n+1}| &=|\lambda x_n+\delta x_{h(n)}|\\ &\geq |\lambda|\,|x_n|-|\delta|\,|x_{h(n)}|\\ &\geq (|\lambda|-|\delta|)|x_n| =q|x_n|. \end{aligned}

Thus |x_n|\geq q^n|x_0|\to\infty, proving instability.

Theorem

Theorem 5

BV39

Let h(n)=\lfloor an\rfloor, where 0<a<1. If

|\delta|>1+|\lambda|,

then the zero solution of (?) is unstable. More precisely, every solution with x_0\ne0 has an unbounded subsequence.

Proof of Theorem 5

Set

c=\frac{|\delta|}{1+|\lambda|}>1.

For every integer k\geq0, the interval

\left[\frac{k}{a},\frac{k+1}{a}\right)

has length 1/a>1. It therefore contains an integer n>k satisfying \lfloor an\rfloor=k.

Starting with k_0=0, suppose that k_j has been chosen, and select n_j>k_j such that \lfloor an_j\rfloor=k_j. From (?),

\delta x_{k_j}=x_{n_j+1}-\lambda x_{n_j}.

Consequently,

\begin{aligned} |\delta|\,|x_{k_j}| &\leq |x_{n_j+1}|+|\lambda|\,|x_{n_j}|\\ &\leq(1+|\lambda|) \max\{|x_{n_j}|,|x_{n_j+1}|\}. \end{aligned}

Choose k_{j+1}\in\{n_j,n_j+1\} so that

|x_{k_{j+1}}|=\max\{|x_{n_j}|,|x_{n_j+1}|\}.

Then k_{j+1}>k_j and

|x_{k_{j+1}}|\geq c|x_{k_j}|.

It follows inductively that

|x_{k_j}|\geq c^j|x_0|\longrightarrow\infty.

Thus the solution has an unbounded subsequence, and the zero solution is unstable.

Source and licence

The text on this page is from arXiv:2610.21635v1 by Prashant M. Gade and Sachin Bhalekar, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.

The record of each item is in analysis/objects/2610/2610.21635.jsonl.