The paper studies a mathematical object called the Luxemburg functional, which is a way of measuring the "size" of a function using a variable exponent. The twist here is that the exponent p(t) is always between 0 and 1, which puts the analysis in what mathematicians call the "concave regime." This is technically much harder than the more commonly studied case where the exponent is greater than 1, because many standard tools from functional analysis break down. The authors carefully build up the right mathematical framework to handle this situation, in particular showing how this functional can be used within a structure called a positive order cone, which is a way of keeping track of functions that are non-negative and organizing them in a useful way.
The authors then apply this framework to a class of nonlocal differential equations. These are equations where the coefficient in front of the second derivative depends not just on the value of the solution at one point, but on an integral involving the solution across the whole interval. This kind of nonlocality makes the equations significantly more complex to analyze. The variable exponent appearing inside that integral is again between 0 and 1, so the Luxemburg functional becomes a natural tool for capturing the behavior of solutions.
The practical payoff of the approach is demonstrated through a concrete numerical example. When you want to prove that solutions to such an equation exist and to pin down roughly where they live in terms of their size, you need good estimates. The authors show that using the Luxemburg functional gives noticeably tighter and more precise bounds on solutions compared to a more standard approach that relies only on the supremum norm, which simply measures the maximum value a function reaches. In other words, their method does not just work in theory; it delivers meaningfully better quantitative information about where solutions can be found.