Maximising the Utility of Virtually Sliced Millimetre-Wave Backhauls via a Deep Learning Approach
Advances in network programmability enable operators to ‘slice’ the physical infrastructure into independent logical networks. By this approach, each network slice aims to accommodate the demands of increasingly diverse services. Precise allocation of resources to slices across future 5G millimetre-wave backhaul networks, so as to optimise their utility, is however challenging. This is because the performance of different services often depends on conflicting requirements, including bandwidth, sensitivity to delay, or the monetary value of the traffic incurred. In this talk, I will present our recent work in which we propose a general rate utility framework for slicing mm-wave backhaul links, which encompasses all known types of service utilities, i.e. logarithmic, sigmoid, polynomial, and linear. We then employ a deep learning solution to tackle the complexity of optimising non-convex objective functions built upon arbitrary combinations of such utilities. Specifically, using a stack of convolutional blocks, our approach can learn correlations between traffic demands and achievable optimal rate assignments. The proposed solution can be trained within minutes, following which it computes rate allocations that match those obtained with state-of-the-art global optimisation algorithms, yet orders of magnitude faster. This confirms applicability to highly dynamic traffic regimes and we demonstrate up to 62% network utility gains over a baseline greedy approach.