Relation between network slices and NFV

Hello all!

I am a beginner when it comes to 5G, NFV, etc. And I have a question regarding network slices: What is a network slice and how is it related to NFV?
Can someone explain it to me in the easiest possible way? :sweat_smile:

Thanks,
Joseph

Hi Joseph,
Interesting question. Let me try to explain the topic my way. So, let’s start with network slicing. Mobile networks traditionally operates (3G, 4G) under a monolithic model in which all services available run within a single “channel”. Since certain services require guaranteed conditions to operate properly, specific mechanisms are employed to provide required Quality of Service (e.g., DiffServ protocol). In 5G, the challenge with QoS is addresed by network slicing. The network slicing concept is based on the idea of every single slice being a logical self-contained and isolated network where services run end-to-end like it would run in its own network. A 5G network would therefore contain multiple slices, usually each of them being optimized for its particular service type, e.g., a video-streaming slice, a slice for IoT, then a slice for critical communications, etc. In a complete slice-aware 5G architecture, slice includes radio network as well as core network resources. Still, concurrent network slices run on a common physical 5G network infrastructure. Most common slices known (also pre-defined) by 5G standardization are eMBB (enhanced mobile broadband), uRLLC (ultra-reliable low latency communications) and mMTC (massive machine-type comm.).
Now, we come to NFV - a virtualization technology over Network Functions which basically enable network slicing, or, to be more concise, they enable management of network slices, i.e., life-cycle management (creating, managing, destroying). By virtualization, Network Functions “transitions” to Virtual Network Functions (VNF) which are decoupled from dedicated hardware and allows them to run on commodity-based servers. Virtualization of Network Functions thus enables upper layer applications to communicate directly to the network: requesting certain network status and monitoring data (e.g., bandwidth available within the slice, QoS status of devices connected, etc.), as well as requesting modifications and/or updates to the network slice (e.g., updating QoS requirements, de-allocating slice, etc.). After all, it still depends on the network infrastructure operator how much of the network functionality would be exposed to the upper layer applications.
Hope this helps clarifying slicing and NFV a bit :wink:
BR, Rudolf.

1 Like