I'm trying to verify that a certain function of two variables $F(x,y)$ satisfies the conditions of a joint CDF. Showing that each condition holds has been fairly straightforward except, that is, for the condition that
$a<b,c<d\implies F(b,d)-F(b,c)-F(a,d)+F(a,c)\geqslant 0$.
I honestly don't know where to start for this one. It's easy enough when the task is to show that this condition is violated but showing that it holds is another matter.
For reference, the particular bivariate function I'm trying to show satisfies this condition is the following (a modification a function Jordan M. Stoyanov examines in Section 5.6 of his book Counterexamples in Probability, 3rd Edition):
Any tips about how I might proceed would be appreciated.


This is also called the 2-increasing property and if your function is continuously twice differentiable, you can equivalently check that
$$\frac{\partial^2 F}{\partial x \partial y} \geq 0 \; .$$
However, your function is not on the transitions from one "block" of definition to another. For those, you can just check as @Satana suggests in the comments by considering all possible combinations of elements from each block. For example, take $0 \leq a < 1$, $1 \leq b < \infty$, $0 \leq c < 1/2$ and $1/2 \leq d < 1$, then
$$F(b,d)-F(b,c)-F(a,d)+F(a,c) = d - c - a/2 + ac \\ = d - c - a(1/2-c) \geq (1/2 - c)(1-a) \geq 0 \text{, etc...}$$