Pantograph Difference Equations: Stability, Multistability, and Chaos

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 $h(i)=\lfloor ai<sup>α\rfloor$, with $0<a<1$ and $0<α<1$. 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 $(λ,δ)$-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.