The paper studies a class of partial differential equations called nonlinear parabolic equations, which describe how quantities like heat or concentration evolve over time and space. The twist here is that the diffusion process, meaning how spreading or smoothing occurs, depends not just on the current state of the system but also on its history. This memory effect is encoded through a mathematical object called a measure-valued kernel, which is a flexible way to assign weights to past times, including the possibility of putting a concentrated chunk of weight at a single moment. The authors carefully separate the contribution at the present instant from the genuinely historical contributions, allowing them to handle both in a unified framework.
The central mathematical results are about well-posedness and stability. Well-posedness means proving that a unique solution exists and behaves reasonably, which requires careful conditions ensuring the diffusion operator is strong enough to control the memory terms. Stability means that if you slightly perturb the kernel that encodes the memory, the solutions change only slightly in response. A particularly subtle scenario the authors handle is when a concentrated weight located at some past time gradually slides forward until it collapses onto the present moment. In that limit, the memory contribution seamlessly merges into the present-time diffusion coefficient rather than causing a discontinuous jump in behavior.
To ground the theory in concrete examples, the authors verify their abstract conditions for a well-known class of equations built around the so-called p-Laplacian operator, which generalizes ordinary diffusion to situations where the spreading rate depends nonlinearly on how steep the gradient is. They consider several specific ways of constructing the memory kernel, including a product form that separates spatial and temporal dependencies, and a regularized family of kernels that approach the boundary case of touching the present time. These examples confirm that the abstract framework captures physically and mathematically meaningful situations, and they illustrate how the general theory can be applied in practice.