The digitization of the cable access network enables the promise of end-to-end network convergence for different lines of service. The promise, however, does not come with a framework and this leads to trepidation on how to proceed. A framework for service convergence must recognize three key principles. The first is that lines of services can have unique prioritization, throughput and latency requirements that need to be met. The second is that within a service there will be distinctions of endpoint types and applications that will eventually need their own unique treatment or policy through the network. The third is that the principles mentioned above are transitional over the life cycle of the service and automation mechanisms to adapt and coexist are necessary to maintain viability over the long term.
In this paper we propose a framework for service convergence using network slicing for MSO networks. We review the network slicing principles for 5G and point to possible analogies that aid in developing a framework for MSO slices including residential services, business services, and mobile services. We cover the concept of network slicing functions which organize and partition network resources available to each service. We describe hard and soft slicing mechanisms, the implementation for slice-aware logical networks and the functionality necessary to maintain end-to-end slice visibility and usability over the life cycle. We also provide several industry recommendations useful for a converged service environment utilizing network slicing, such as open interfaces for core functions, QoS implementation in packet networks, listing of slice expectations, usability of hard and soft slicing.