O-RAN PlugFest Front Haul Transport Validation

0:03 welcome my name is Christoph Sarkis from

0:05 Juniper Networks I would like to present

0:08 the join the testing performer

0:10 telefonica Juniper Networks and keysight

0:12 as part of the Orient lock Fest 2022

0:16 transport testing so why Oran frontal

0:19 testing is important well because we

0:21 need to provide a good solution to the

0:25 customers and I would like to ask Luis

0:27 from telephonica as well to express his

0:30 view on this please yeah actually this

0:33 is very relevant to understand the

0:34 behavior of the of the equipment on the

0:36 lab essentially the idea is to yeah to

0:39 have a clear

0:40 overview of what could be the the

0:42 respective Behavior once we move this

0:44 into the field and this is clearly

0:46 interesting for operations but not only

0:48 for operations also for planning for

0:50 engineering the whole process of moving

0:52 the equipment and implementing the

0:54 Solutions in the field great so Juniper

0:57 has the full portfolio to address the 5G

0:59 mobile transport and it's starting from

1:01 the small ACX 7024 which can be used on

1:04 the cell site router so on the on the

1:06 sidestart locations going through the

1:09 ACX 7348 ACX 7509 up to 86 7508 which is

1:15 the big box that can be used in the

1:16 aggregation site

1:18 and you know let's discuss the test

1:20 setup that we have used during the

1:22 testing so basically we have small

1:25 frontal network with two cell side

1:27 routers which are Juniper AC extorters

1:29 and one halfside router which is as well

1:31 Juniper ACX rotor and to these two small

1:35 Network we have Ariel du and CU

1:38 emulators emulated by keysight connected

1:40 and different locations and of course we

1:43 have as well some background traffic to

1:45 saturate the links and as well a grand

1:48 master emulated emulated by keysight so

1:52 let me give now the words to avik from

1:55 his sites to explain more data as about

1:57 the testing okay thank you Kristoff the

2:00 online transport is you know very

2:02 different from a traditional transport

2:03 in the sense that it needs to carry time

2:05 sensitive Radio Data through the packet

2:07 switch frontal so validation of such a

2:10 frontal transport requires creating

2:12 realistic traffic pattern for the cus

2:15 and in plain messages the timing pattern

2:18 and all of these with large number of

2:20 deals are use emulation so key sites are

2:23 you doing emulation capability based on

2:25 IX Network and novas Hardware allows the

2:28 user to measure transport performance at

2:30 high scale it validates the kpis that is

2:33 required to carry the latency sensitive

2:35 Radio Data and therefore helps test

2:38 compliance as per the order an alliance

2:40 wg9 test specification and in a system

2:43 level test the solution can be used to

2:46 identify potential bottlenecks at

2:48 specific transport elements for example

2:50 the transceivers switches smartness

2:53 operating systems or even the container

2:55 engines

2:56 while validating the overall transport

2:59 infrastructure at keysight we look at

3:01 the six technology enablers for us the

3:05 six pillars of 5G ran transport are the

3:07 time synchronization the deterministic

3:09 nature of traffic the encapsulations

3:12 that carry the radio data over Ethernet

3:14 or ECP in other words the network

3:18 slicing enabled by Technologies like

3:20 segment routing Etc the virtual

3:22 infrastructure that hosts the virtual

3:25 RAM and you know last but not the least

3:27 the security and encryption aspects so

3:31 our product portfolio spans across

3:33 different stages of the life cycle

3:35 starting from pre-silicon validation at

3:37 the left hand side which is covered by

3:40 the product called likes verify then IX

3:43 Network along with Nova's Hardware

3:44 covering the infrastructure test the

3:47 conformance tests through IX Anvil the

3:50 network impairment test through our

3:52 brand new product called any three and

3:54 finally the virtual infrastructure

3:56 validation through Cloud Peak

3:59 now coming to the specific test

4:01 configurations used for this plug Fest

4:04 each of the emulated ODU Oru pair is

4:07 configured with 100 megahertz downlink

4:10 carrier and 100 megahertz Uplink carrier

4:13 with 30 kilohertz subcarrier spacing

4:16 this results to 1.75 gbps of ECP traffic

4:20 stream in both Uplink and downlink

4:23 direction

4:24 the traffic streams between ocu and ODU

4:27 is configured with 1.2 gigabits of you

4:30 know throughput in both Direction and

4:33 also background traffic streams

4:35 traversing the HSR or the hopsight

4:37 router and each of the csrs or the cell

4:40 site routers is configured with 38.2

4:43 gbps of throughput in both Direction

4:46 the ECP streams are configured with

4:48 proper cos settings for strict priority

4:52 queuing so that they experience no loss

4:55 and minimum Network delay

4:58 with this configuration traffic is

5:00 blasted through the front hall to

5:03 measure loss latency and packet delay

5:05 variation no packet loss was observed

5:08 during the test as shown in the IX

5:11 Network statistics

5:12 also maximum latency for EC pre-streams

5:16 was observed as about 20 microseconds

5:19 minimum was about 5 microseconds and

5:22 average was about 6 microseconds The IX

5:26 network based clock quality analysis

5:28 tool was used to measure explain

5:31 characteristics average time error on

5:34 emulated ODU ports that was running a

5:37 PTP follower clock was about one

5:39 nanosecond it means the HSR added only

5:42 one nanosecond of constant time error

5:45 while propagating time to the ODU

5:48 on the audio side the average time error

5:51 observed on each emulated Oru ports was

5:54 about 75 nanoseconds maximum relative

5:58 time error turned out to be about 10

6:01 nanosecond at Oru Unis of those orus

6:05 so now I will hand it over to Luis once

6:09 again to let us know about the

6:11 importance of this validation

6:13 so thanks avik so we have seen uh along

6:17 the presentation there is a diverse and

6:20 Rich number of options in the deployment

6:22 of from four mutual and Bachelor

6:24 Solutions so this is the this shows the

6:27 relevance of having a clear

6:29 characterization of the solutions before

6:30 going into the field

6:32 not only from the operational point of

6:33 view but also the engineering the

6:35 planning all of this is a

6:38 is is impacted let's say by a good

6:41 characterization before going into the

6:43 field so

6:44 um I would like to thank unipan and

6:46 Chrisette for this exercise it was a

6:48 very uh insightful and relevant for us

6:51 and yeah uh thanks for for participating

6:54 on on this list