The Algorithmic Aspects of Network Slicing

Abstract

Network slicing is a technique for flexible resource provisioning in future wireless networks. With the powerful SDN and NFV technologies available, network slices can be quickly deployed and centrally managed, leading to simplified management, better resource utilization, and cost efficiency by commoditization of resources. Departing from the one-type-fits-all design philosophy, future wireless networks will employ the network slicing methodology in order to accommodate applications with widely diverse requirements over the same physical network. On the other hand, deciding how to efficiently allocate, manage and control the slice resources in real-time is very challenging. This paper focuses on the algorithmic challenges that emerge in efficient network slicing, necessitating novel techniques from the communities of operation research, networking, and computer science.

Orchestration and Control in Software‐Defined 5G Networks: Research Challenges

“…This has many causes: for instance, (i) the control and management activity is moved to separate platforms that must also be managed, see, e.g., [42,43]; (ii) the system is in general multivendo; (iii) implementing the improved functionalities of the M&O causes additional complexity, etc. ; (iv) efficient allocation, management, and control of assigned resources in real-time are presented, for instance, in [44]. Finally, continuous dynamic adaptation to changing workloads, new services, failures, and repairs simultaneously which provide service continuity amplifies the challenges.…”

Content-oriented network slicing optimization based on cache-enabled hybrid radio access network

“…The factors affecting the hit rate of the system include the average degree value of the graph, the connectivity feature of the graph (ER model or BA model), the scale of the network (the number of nodes), the number of copies of the contents caching in the network, and the weight value for each hop in/ between different level as well as the popularity of the contents and caching policy of each node in the network. With the formulation (10)(11)(12)(13)(14)(15), if the hit rate in the UE level is greater, then the system cost E c is smaller, in other words, it is better that the contents can be hit in the UE level. However, contents cannot be cached in every UE because of the limited caching space in UE, the challenge is to find the tradeoff between the hit rate and the system cost.…”

“…Orchestration at the operator level can be seen as a special instance of the virtual network embedding (VNE) problem, in which the combined network slicing optimization problem is to make decision on optimal placement of VNFs at resource nodes, jointly the necessary link capacity reservations for their interconnection, under additive link and node capacity constraints so as to minimize the overall resource utilization cost [14]. The typical slicing motivation are different operators, different infrastructures and resource level, as well as various service requirements, including low power, low-data-rate machine-type communication, high data rate multimedia, and delaysensitive applications [8].…”

“…The typical slicing motivation are different operators, different infrastructures and resource level, as well as various service requirements, including low power, low-data-rate machine-type communication, high data rate multimedia, and delaysensitive applications [8]. Among the network slicing issues, network slicing based on content-oriented service become a challenging research direction as one of the important services provided by heterogeneous radio access network [7,14].…”

“…Some optimization issues are designed from the point of view of VNF placement [14,30,31]. In [30], a VNF placement heuristic named wireless network embedding (WiNE) is presented to optimize the number of accepted requests, the average embedding cost, the average node, and link utilization with k-ary fat-tree substrate network.…”