I think you are confused about what you can do with MATLAB. Those constraints simply tell you the associated matrix for the graph is a symmetric one, and places bounds on the elements of that graph. For example, symmetry tells you the "cost" to move from node i to node j is the same as the cost to move in the opposite direction.
But you cannot just create a general unspecified matrix that has those properties built into the matrix definition, and then go on to make some conclusions about the result.
If you had a specific numerical instance of a graph, you could test and verify it satisfies those constraints. (If you knew the value of Umax, that is.) But that is different from building those properties into the matrix itself, and then using MATLAB to come to some conclusion, valid over the set of all such graphs.
