Joint Resource Optimization, Computation Offloading and Resource Slicing for Multi-Edge Traffic-Cognitive Networks