What are some unanswered questions in networking that, if solved, will revolutionise the field?

For example, math and computer science have quite a few open problems. Is there such a thing in networking?