Open RAN Service Management and Orchestration & RAN Intelligent Controller Demo

0:00 [Music]

0:01 hi everybody I am nataraj and

0:03 krishnamurthy product marketing manager

0:05 for Juniper open range Solutions and

0:08 today I am going to talk about how

0:10 service providers can achieve new levels

0:13 of service agility Automation and

0:15 operational efficiency by unlocking

0:17 Innovation and value in the ram so we

0:20 are going to talk about two Juniper

0:22 Solutions Juniper ran intelligent

0:23 controller and Orion service management

0:25 and orchestration so let me start with

0:27 juniper Ryan intelligent controller I am

0:30 happy to share that Juniper Ryan

0:31 intelligent controller has been awarded

0:33 the glomo award for best network

0:36 software breakthrough at Mobile World

0:37 Congress 2023. this is in partnership

0:40 with rakutan Symphony for the record and

0:42 Sim World platform with juniper Ryan

0:44 intention controller Juniper provides a

0:46 ran intelligent controller platform

0:48 which is fully compliant with uran

0:50 Alliance specifications and interfaces

0:53 we have the near real time rig we have

0:56 the non-real time rig as well as the X

0:58 apps and the RX

1:00 platform also provides an open API and

1:03 an SDK to onboard applications from

1:06 third-party vendors this gives greater

1:08 flexibility and choice of suppliers to

1:11 service providers in this demo we are

1:13 showcasing three different use cases

1:15 admission control traffic steering and

1:17 Energy savings

1:18 the admission control and traffic

1:20 steering use cases were demonstrated

1:22 with Vodafone on a live Network running

1:25 commercial traffic parallel Wallace was

1:27 providing the rent for that deployment

1:30 so let me get into this admission

1:32 control use case admission control is

1:34 used to deliver Priority Services for

1:35 any user type at any time for example it

1:39 could be for Public Safety users schools

1:41 hospitals Etc this is the setup that we

1:44 use to kind of validate this use case so

1:46 parallel Wireless was providing the rent

1:48 we had the Juniper near real time brick

1:49 non-real time brick and Juniper was

1:52 providing the admission control RF and X

1:54 app traffic steering RF was coming from

1:56 juniper and traffic steering except was

1:58 coming from parallel Wireless this is an

2:01 example of how a third party vendor has

2:03 developed plan except on top of juniper

2:05 near real time brick platform so let's

2:08 get into this use case so we have a

2:10 landscape with a suburban area and a

2:13 rural area

2:13 so let's start with by configuring

2:16 admission controller so we are first

2:18 starting with juniper non-real time

2:19 brick where we are going to onboard the

2:22 admission control RM the next step is we

2:25 have two near RT rigs where we are going

2:28 to onboard admission control accepts so

2:30 here you could see two admission control

2:32 accepts running one in the rural area

2:34 one in the Suburban area

2:36 the tenant we have here is a public

2:39 safety tenant he has subscribed for 2000

2:42 us as SLA by default 1000 have been

2:46 allocated to the suburban area thousand

2:48 has been allocated to the rural area we

2:50 have a normal situation here there is no

2:52 emergency the number of active duties is

2:55 well within this thousand limit so you

2:57 don't see any rejections here now let's

2:59 say we have an emergency uh in the rural

3:01 area there is a fire in the rural area

3:03 you have First Responders firefighters

3:06 Healthcare Personnel arriving in the

3:09 rural area and trying to connect to the

3:10 network and you suddenly see a kind of a

3:13 strike in the number of active ues in

3:15 the rural area you have about 242

3:17 rejections for this tenant so this is

3:19 not a situation that we want so this is

3:21 where Rick comes into picture Rick can

3:23 get triggered by an external enrichment

3:25 server or it can identify on its own

3:28 that there are rejections in the rural

3:30 area and automatically increase the

3:32 allocation of resources in the rural

3:34 area

3:35 here you could see the trick has

3:37 increased the allocation in the rural

3:38 area from 1000 to 1800 it has reduced

3:41 allocation in suburb from 1000 to 200

3:43 and you can see here that the number of

3:45 rejections has come in the rural area

3:47 has come down to zero

3:50 and you could see that on the in the

3:52 rural area all the active ues who needed

3:54 connections are getting connections

3:56 this is a simple deployment where we are

3:59 showing a single tenant but in a typical

4:02 deployment you could have hundreds of

4:03 such tenants each of them having needing

4:06 their own resources and that's why you

4:09 need something like a rig to be able to

4:10 manage this now the emergency situation

4:12 has gone away

4:13 so the allocation has gone back to a

4:17 thousand each in Suburban rural area so

4:19 in summary admission control is able to

4:21 provide Dynamic allocation of network

4:23 resources without having to over

4:24 provisioning resources at each location

4:26 and the services here are automatically

4:28 optimized by the ram based on real-time

4:31 data since this is based on open run if

4:33 you develop an application once it can

4:35 be used across all vendors without

4:37 having any vendor specific

4:38 implementations you will take a quick

4:40 look at the oran service management

4:42 orchestration

4:43 before we get into the demo let's look

4:45 at why do you need a new approach to

4:47 service management and orchestration the

4:50 first is evolving architectures service

4:52 providers are moving from closed

4:54 monolithic solutions to virtualized

4:56 disaggregated and open Solutions second

4:59 massive scale with 5G and orange

5:01 services are going to be delivered from

5:03 thousands of clouds and tens and

5:06 thousands of radio sites the third is

5:09 drive for on-demand service models

5:11 service providers want to spin up new

5:13 custom services on demand and deliver

5:16 them under SLS and charge them on a as a

5:19 service basis to overcome this

5:21 complexity Juniper has is providing a

5:23 service management orchestration

5:24 platform for open Ram its fully standard

5:28 compliant supporting 3gpp TMF and open

5:30 run standards it implements the Ryan

5:32 Network slice management function it has

5:35 the near real time brake non-real time

5:36 brick it has 4com which is Federated Oak

5:39 load management and orchestration which

5:41 takes care of management of

5:43 infrastructure resources like data

5:45 centers clusters on top of which the

5:48 network functions are going to be

5:49 deployed it has the network function

5:51 orchestrator the nf4 which takes care of

5:54 orchestrating the network functions such

5:56 as ODU ocu near real time brick on top

5:59 of this cloud

6:01 so we are going to demonstrate Oran

6:03 smo's capabilities with this topology

6:05 which we have a Barcelona landscape

6:07 where we have a regional Cloud three

6:09 different Edge clouds along with their

6:11 orus

6:13 we are going to show how Quran Smo can

6:15 be used to deploy and manage slices to

6:18 an Enterprise as well as an industry we

6:21 will be walking through the entire life

6:22 cycle of a network slice starting with

6:25 the preparation phase where we are going

6:27 to onboard Cloud infrastructure designs

6:29 license service templates reduce service

6:32 creation do service activation do

6:34 service operations to ensure the slas

6:36 within the confirmed limits and as well

6:39 as flash termination Juniper Smo has

6:42 been designed with a clean slate

6:43 architecture to implement and monetize

6:46 Network slicing it uses open interfaces

6:49 across the entire architecture to

6:51 support multi-vendor purposes