A general method for constructing high-dimensional chaotic maps with topological mixing on the global phase space

High-dimensional chaotic maps offer a larger parameter space, increased complexity, and enhanced resilience against dynamical degradation compared to their one-dimensional counterparts. Therefore, they are gradually replacing one-dimensional chaotic maps in various applications. However, many methods for generating high-dimensional chaotic maps lack mathematical proofs, which cannot theoretically ensure their chaotic nature. Even high-dimensional chaotic maps with theoretical support often lack global transitivity and exhibit local chaos. Applying such chaotic maps in chaos-based stream ciphers or random number generators results in poor randomness of generated chaotic sequences, reduced internal state space, and numerous weak keys, which is not ideal. This paper proposes a systematic method for constructing high-dimensional chaotic maps (called dispersal maps). The paper proves that the maps constructed are topologically mixing across the entire space and are hyper-chaotic on an invariant subset of full measure. These properties make them satisfy almost all definitions of chaos, and their chaotic dynamical behavior is global: exhibiting transitivity across the entire phase space rather than a local subregion, a dense scrambled subset rather than a tiny one, and being hyper-chaotic almost everywhere rather than on a local attractor. Therefore, dispersal maps can improve the existing problems of locally chaotic maps in application. The experiments also indicate that dispersal maps exhibit ergodicity on the phase space, with highly uniform trajectory distributions and sensitivity to initial perturbations. The findings provide researchers with ideal chaotic maps and a feasible method for constructing high-dimensional chaotic maps with global chaos.