5G
NFV MEC Kubernetes
1
Module : NFV,MEC and K8 in 5G
• Claudication/Softwarization of elements and benefits
• Virtualization Introduction
• Benefits
• Network Functions Virtualization
• MANO
• Need of NFV and SDN are used in 5G
• MEC Introduction
• Mobile Edge computing Introduction (MEC)
• MEC in 5G
• Benefits of MEC
• MEC Platform
• MEC Use cases
• Kubernetes Overview
• K8 in 5G
2
Softwarization
• Softwarization
and virtualization are two
paradigms considered to be at the basis of
the design process of 5G networks
• Benefits:
• Provides
provide flexibility, configurability in
heterogeneous uses cases of 5G network
3
Virtualization
4
Network Function Virtualization
Network functions virtualization (NFV) is an initiative to virtualize network
services traditionally run on proprietary, dedicated hardware. With NFV,
functions like routing, load balancing and firewalls are packaged as virtual
machines (VMs) on commodity hardware. Individual virtual network functions,
or VNFs, are an essential component of NFV architecture.
5
NFV Reference Architectural
Framework
Os-Ma
Or-Vnfm
Ve-Vnfm
Vi-Vnfm
Nf-Vi
6
Practice in NFV
• Separation of software from hardware
• Flexible deployment of network functions:
• Dynamic service provisioning
7
MANO E2E Network Service
Logical Network Service
Abstractions
VNF PNF
End End
Point VNF Point
VNF VNF
Logical Links
VNF Instances
SW Instances
VNF VNF VNF VNF Management
and
Orchestration
NFV Infrastructure (NFVI)
Virtual Resources vCompute vStorage vNetwork
Virtualization SW Virtualization Layer
HW Resources Compute Storage Network
PNF: Physical Network Function
VNF: Virtualized Network Function
8
Key Functional Nodes in MANO
• NFV Orchestrator (NFVO)
• VNF Manager (VNFM)
• Virtualised Infrastructure Manager (VIM)
NFV management and orchestration functions can be grouped in three broad
categories: virtualised resources, virtualised network functions, and network
services.
9
SDN
The core similarity between software-defined networking (SDN) and network
functions virtualization (NFV) is that they both use network abstraction. SDN
seeks to separate network control functions from network forwarding
functions, while NFV seeks to abstract network forwarding and other
networking functions from the hardware on which it runs. Thus, both depend
heavily on virtualization to enable network design and infrastructure to be
abstracted in software and then implemented by underlying software across
hardware platforms and device
10
MEC (MOBILE EDGE
COMPUTING)
11
Cloud-native architecture
BTS Core
Today • Current radio is distributed
• Current core is centralized
Large number Very few
Edge cloud
BTS Radio Core Core • Radio more centralized for
5G processing processing faster scalability
System
• Core more distributed for low
latency
Cloud-native architecture to deliver massive scalability, performance, flexibility and reliability to meet the
economics of IoT/MTC and broadband evolution, and a foundation for 5G
12
Drivers of MEC
• Low Latency
• Increases speeds
• Data Analytics
13
Use Cases of MEC
• Video analytics
• location services
• Internet-of-Things (IoT)
• Augmented reality
14
Mobile Edge Computing framework
mobile edge host, including the following:
- mobile edge platform;
- mobile edge applications;
- virtualization infrastructure
Mobile edge host is an entity that contains a mobile edge platform and a
virtualization infrastructure which provides compute, storage, and network
resources, for the purpose of running mobile edge applications.
Mobile edge platform is the collection of essential functionality required to
run mobile edge applications on a particular virtualization infrastructure and
enable them to provide and consume mobile edge services. The mobile edge
platform can also provide services.
Mobile edge applications are instantiated on the virtualization infrastructure
of the mobile edge host based on configuration or requests validated by the
mobile edge management
15
MEC “ETSI System Reference Architecture”
ME platform is deployed as a VNF.
The virtualization infrastructure is deployed as a NFVI
Reference points regarding the mobile edge platform functionality (Mp)
management reference points (Mm)
reference points connecting to external entities (Mx)
16
5G and MEC system architecture
Figure 1 Figure 2
MEC application, i.e. an application hosted in the distributed cloud of a MEC
system can belong to one or more network slices that have been configured in
the 5G core network.
Policies and rules in the 5G system are handled by the PCF. The PCF is also
the function whose services an AF, such as a MEC platform
17
MEC Host in 5G(1)
MEC platform would leverage the 5G network architecture and performs the
traffic routing and steering function in the UPF. For example, a UL Classifier of
UPF is used to divert to the local data plane the user traffic matching traffic
filters controlled by the SMF, and further steer to the application. The PCF and
the SMF can set the policy to influence such traffic routing in the UPF. Also
the AF via the PCF can influence the traffic routing and steering
18
MEC Flow In 5G
19
Capabilities exposure
5G capability exposure to the MEC system. In this case the MEC orchestrator
(MEC system level management) appears as a 5G AF, providing centralized
functions for managing the computing resources and operation of the MEC
hosts. In addition, it offers orchestration of MEC applications running on MEC
hosts. The MEC orchestrator as a 5G AF interacts with NEF and with other
relevant NFs with regards to overall Monitoring, Provisioning, Policy and
Charging capabilities. The MEC host, on the other hand, might be deployed at
the edge of 5G RAN to leverage the advantages of MEC for optimizing the
performance of applications and improving users’ Quality of Experience.
Therefore, it is possible that the MEC platform may need direct exposure to
the Centralized Units (CUs) of the 5G RAN and potentially even the
Distributed Units (DUs)
20
Migration pattern of MEC
Deployment
21
MEC Deployment Scenarios in 5G
1.MEC and the local UPF collocated with the Base Station.
2.MEC collocated with a transmission node, possibly with a local UPF
3.MEC and the local UPF collocated with a network aggregation point
4.MEC collocated with the Core Network functions (i.e. in the same data
centre)
22
Integrated MEC deployment in 5G
network
MEC orchestrator is a MEC system level functional entity that, acting as an
AF, can interact with the Network Exposure Function (NEF),
The User Plane Function (UPF) has a key role in an integrated MEC
deployment in a 5G network. UPFs can be seen as a distributed and
configurable data plane from the MEC system perspective. The control of that
data plane, i.e. the traffic rules configuration, now follows the NEF-PCF-SMF
route. Consequently, in some specific deployments the local UPF may even
be part of the MEC implementation.
In the MEC system on the right-hand side of Figure the MEC orchestrator is a
MEC system level functional entity that, acting as an AF, can interact with the
Network Exposure Function (NEF),
. On the MEC host level it is the MEC platform that can interact with these 5G
NFs, again in the role of an AF. The MEC host, i.e. the host level functional
entities, are most often deployed in a data network in the 5G system. While
the NEF as a Core Network function is a system level entity deployed
centrally together with similar NFs, an instance of NEF can also be deployed
in the edge to allow low latency, high throughput service access from a MEC
host
23
MEC use cases
24
Video Optimization
RAN-AWARE Video Optimization: Currently, mobile video streaming
capabilities may suffer from sluggish video buffering times. Video buffering is
caused by the Transmission Control Protocol (TCP) not adapting fast enough
to varying radio conditions. MEC technology will dodge those video streaming
issues by communicating to the video server the best bit rate for the given
radio conditions, which reduces the buffering time of the device’s video
stream. See the above graphic for more info on how this works.
25
Intelligent Video Acceleration
26
Local Content cahcing
27
Location based services
28
Application computation offloading
29
Ref Achitecture
30
Micro MDC
31
C-RAN Architecture
32
Deployment Evolution
DISCLAIMER
33
Virtual RAN Deployment
34
KUBERNETES(K8)
35
Introduction
• Kubernetes(K8s) is an open-source system for
automating deployment, scaling, and
management of containerized applications.
• Googleinitially and now Cloud Native
Computing Foundation.
• Kubernetes can be deployed in platform or
infrastructure as a service (PaaS or IaaS)
• It
works with a range of container tools i.e.
Dockers
K8 is open-source container orchestration system for
automating application deployment, scaling, and
management
36
Benefits
• Provides required resources Quickly and
efficiently
• Kubernetes is cost efficient.
• Kubernetes is portable.
• Runs on Amazon
WebServices (AWS), Microsoft Azure, and
the Google Cloud Platform (GCP)
37
Dockers
Tool to run application in ISOLATED Environment
Docker is a computer program that performs operating-system-level
virtualization
38
CONATINERS
Running image of an environment
Containers are Micro services
•OS
•Software
•App
Reside inside POD
39
VM vs Docker
Virtual Machines are slow and take a lot of time to boot.
Containers are fast and boots quickly as it uses host operating system and
shares the relevant libraries.
Containers do not waste or block host resources unlike virtual machines.
Containers have isolated libraries and binaries specific to the application they
are running
40
Kubernetes Objects
• PODS
• SERVICES
• VOLUMES
• NAMESPACES
PODS
The basic scheduling unit in Kubernetes .
It adds a higher level of abstraction by grouping containerized components. A
pod consists of one or more containers that are guaranteed to be co-located
on the host machine and can share resource.
SERVICES
Kubernetes service is a set of pods that work together, such as one tier of
a multi-tier application.
VOLUMES
Filesystems in the Kubernetes container provide ephemeral storage, by
default. This means that a restart of the container will wipe out any data on
such containers, and therefore, this form of storage is quite limiting in
anything but trivial applications.
NAMESPACES
Kubernetes provides a partitioning of the resources it manages into non-
overlapping sets called namespaces. They are intended for use in
environments with many users spread across multiple teams, or projects, or
41
even separating environments like development, test, and production.
41
Kubernetes cluster
42
K8 Architecture
etcd: etcd is a persistent, lightweight, distributed, key-value data store developed by CoreOS that
reliably stores the configuration data of the cluster, representing the overall state of the cluster at any
given point of time. Just like Apache ZooKeeper, etcd is a system that favors Consistency over
Availability in the event of a network partition (see CAP theorem). This consistency is crucial for
correctly scheduling and operating services. The Kubernetes API Server uses etcd's watch API to
monitor the cluster and roll out critical configuration changes or simply restore any divergences of the
state of the cluster, back to what was declared by the deployer. As an example, if the deployer specified
that three instances of a particular pod need to be running, this fact is stored in etcd. If it is found that
only two instances are running, this delta will be detected by comparison with etcd data, and
Kubernetes will use this to schedule the creation of an additional instance of that pod.
API server: The API server is a key component and serves the Kubernetes API using JSON over HTTP,
which provides both the internal and external interface to Kubernetes. The API server processes and
validates REST requests and updates state of the API objects in etcd, thereby allowing clients to
configure workloads and containers across Worker nodes.
Scheduler: The scheduler is the pluggable component that selects which node an unscheduled pod (the
basic entity managed by the scheduler) runs on, based on resource availability. Scheduler tracks
resource use on each node to ensure that workload is not scheduled in excess of available resources.
For this purpose, the scheduler must know the resource requirements, resource availability, and other
user-provided constraints and policy directives such as quality-of-service, affinity/anti-affinity
requirements, data locality, and so on. In essence, the scheduler's role is to match resource "supply" to
workload "demand".
Controller manager: A controller is a reconciliation loop that drives actual cluster state toward the
desired cluster state. It does this by managing a set of controllers. One kind of controller is a replication
controller, which handles replication and scaling by running a specified number of copies of a pod
across the cluster. It also handles creating replacement pods if the underlying node fails.Other
controllers that are part of the core Kubernetes system include a "DaemonSet Controller" for running
exactly one pod on every machine (or some subset of machines), and a "Job Controller" for running
pods that run to completion, e.g. as part of a batch job. The set of pods that a controller manages is
43
determined by label selectors that are part of the controller's definition.[
The controller manager is a process that runs core Kubernetes controllers like DaemonSet Controller
and Replication Controller. The controllers communicate with the API server to create, update, and
delete the resources they manage (pods, service endpoints, etc.)
43
