Complex dynamics in two-dimensional coupling of quadratic maps
Anca Radulescu, Eva Kaslik, Alexandru Fikl
TL;DR
This work extends the classical Mandelbrot framework to two-dimensional Complex Quadratic Networks by introducing and analyzing the equi-M set M_2, a postcritically bounded locus for coupled quadratic maps. The authors develop a rigorous 2D-CQN model, derive the escape radius and stable fixed-point conditions, and characterize the main equi-cardioid via Schur–Cohn stability criteria, complemented by explicit analyses of feed-forward and equal-row-sum subfamilies. Key findings show that, unlike the univariate case, higher-period attractors and coexisting dynamics decouple the critical orbit from the full combinatorial dynamics, causing the equi-M set to be only a partial atlas of behavior; synchronization phenomena are also clarified, with coupling often forcing M-synchronization under broad conditions. Overall, the results reveal richer, more intricate dynamics in 2D CQNs and establish a bottom-up foundation for understanding how coupling shapes the dynamics of larger, high-dimensional networks.
Abstract
This paper examines the structure and limitations of equi-M sets in two-dimensional Complex Quadratic Networks (CQNs). In particular, we aim to describe the relationship between the equi-M set and the parameter domains where the critical orbit converges to periodic attractors (pseudo-bulbs). The two-node case serves as a foundational testbed: its analytical tractability enables the identification of critical phenomena and their dependence on coupling, while offering insight into more general principles. The two-node case is also simple enough to allow for explicit coupling conditions that govern phase transitions between synchronized and desynchronized behavior. Using a combination of analytical and numerical methods, the study reveals that while the period-1 pseudo-bulb closely tracks the boundary of the equi-M set near its main cusp, this correspondence breaks down for higher periods and in regions supporting coexisting attractors. These discrepancies highlight key differences between single-map and coupled dynamics, where equi-M sets no longer provide a full encoding of system combinatorics. These findings clarify the topological and dynamical behavior of low-dimensional CQNs and point toward a sharp increase in complexity as the number of nodes grows, laying the groundwork for future studies of high-dimensional dynamics.
