Early lessons from Open RAN Deployment in brownfield ,a must delivery model to untap 5G Scale,Complexity and economy in 2021+

With most of Tier1 Operators rolling the ball for early experience of 5G Standalone based on 3GPP Release-16 that can offer very new and unique experiences to 5G  around uRLLC and mIOT along with improved QOS for broadband the whole industry is looking forward to accelerate 5G adoption in 2021 .

This is an ideal time for the industry to find new ways to improve human life and economy in the post covid world . However the biggest problem with 5G experience so far has been the Delivery model that can offer both Cost and Scale advantages . With the hyperscalers eyeing the Telco billion dollar market it is necessary for all vendors and Telco’s themselves to analyze and find ways towards new business models that are based on

  • Coherence
  • Data Driven
  • Service Quality

By coherence I mean when we go to 5G Radio site what components will be residing there

  • A 5G Radio site components
  • Edge Applications
  • Network connectivity

By Data I mean till now depsite dleiverying thouand’s of 5G sites we can not offer DaaS(Data as a service) to vertical’s the only visibiltiy we have is on horizontal interfaces in a 3GPP way .

The third and the most important piece is RF and RAN service . You talk to any RAN guy and he is not interested in both Coherence and Data unless we answer him a RAN service that is atleast same lr better the legacy .

This makes the story of Open RAN very exciting to analyze and this is the  story of my experience of leading such projects during last years in both Company and Industry . This is my humble opinion that Open RAN and other such innovative solutions must not be analyzed only through tehnology but an end to end view where some use cases are specificcally requiring such solutions to be succcessful .

Why Open RAN

For me the Open RAN is not about Cloud but finding a new and disruptive Delivery model to offer RaaS (RAN as a service). Have you ever imagined what will happen if a Hyperscaler like AWS or Azure acquire a couple of RAN companies who can build a Cloud Native RAN Applications  , the Telco can order it online and it can be spined in the Data center or on the Edge Device in a PnP(Plug and Play manner)

If you think I am exaggerating  already there are disucssions and PoC’s happening aorund this . So Open RAN is not about Cloud but about doing somethnig similar by the Telco Industry  in a Open and Carrier grade fashion .

This is where Open RAN is gaining momentum to bring power of Open , Data driven and Cloud based disaggregated solution to the RAN site .Future of Telco is a well-designed software stack that extends all the way from Core the last mile Site of Network  Crucially, it also allows for placing more compute, network and storage closer to the source of unrelenting volume of data – devices, applications, and end-users.

There is another aspect which is often overlooked which is transport cost , as per filed trial result use of Open RAN CPRI interface 7-2 increased the front haulf capacity utilziation by at least 30-40% primarily as there are lot of propietry overheads in CPRI world .

“For CXO it means  at least 30-40% direct savings on Metro and transport Costs”

What is Open RAN

5G RAN has a huge capacity requirements and naturally to scale this network disaggregation and layered network architecture is the key . Open RAN architecture can be depicted as follows .

RU:

Radio Unit that handles the digital front end (DFE) and the parts of the PHY layer, as well as the digital beamforming functionality .It is almost same architecture as in DBS (Distributed Basestation) architecture fofered by many legacy vendors . Conceptually AAU (Active antenna unit) is considered together with radio unit

DU:

Distributed unit handles the real time L1 and L2 scheduling functions mainly MAC split . it is realt time part of BBU (Baseband unit)

CU:

responsible for non-real time, higher L2 and L3 , it is non real time functions of BBU like reosurce pooling , optimization etc

RIC:

RAN Intelligent contorller is the intelligence component of Open RAN that will collect all data and offer insights and innovation through xAPPS . similarly it cooperates with NT-RIC (Non real time RIC) which is part of ONAP/MANO to offer end to end end SMO (Service management and Orchestration) functions  

Interfaces of Open RAN

When it comes to understand Open RAN we need to understand both those defined by O-RAN and those by 3GPP and this is so important as it do require cross SDO liaisonship

O-RAN interfaces

  • A1 interface is the inerface between non real time RIC /Orchestrator and RAN components
  • E2 interface is the interface between NT-RIC (RAN Intelligent Controller) and CU/DU
  • Open FrontHaul is the interface between RU and DU mostly we are focusing on eCPRI 7-2 to standardize it
  • O2 interface is the interfaces betwee NFVI/CISM and Orchestrator

3GPP interfaces

  • E1 interface  is the the interface between CU-CP (control plane) and CU-UP (user plane)
  • F1 interface is the interface between CU and DU
  • NG-c is the interface between gNB CU-UP and AMF in 5G Core Network

To solve all interface and use case issues the ORAN Alliance is working in a number of streams  to solve issues .

  • WG1: Use Cases and Overall Architecture
  • WG2: The Non-real-time RAN Intelligent Controller and A1 Interface
  • WG3: The Near-real-time RIC and E2 Interface
  • WG4: The Open Fronthaul Interfaces, 
  • WG5: The Open F1/W1/E1/X2/Xn Interface
  • WG6: The Cloudification and Orchestration
  • WG7: The White-box Hardware Workgroup
  • WG8: Stack Reference Design
  • WG9: Open X-haul Transport.
  • Standard Development Focus Group (SDFG): Strategizes standardization effort. Coordinates and liaises with other standard organizations.
  • Test & Integration Focus Group (TIGF): Defines test and integration specifications across workgroups.
  • Open Source Focus Group (OSFG): Successfully established O-RAN SC to bring developer in the Open RAN ecosystem

Early Playground of Open RAN

The change in world economy impacted by geo political factors like a drive to replace Chinese vendors from networks like in Australia for national security reasons naturally change momentum to find both less costly and high-performance systems. Naturally one of the prime networks where Open RAN will get introduced are above .It is true that still there are some gaps in Open RAN performance mainly on the Base band processing and front haul but there are some use cases in which Open RAN already proved to be successful as shown below , the key point here is that although there are some issues but with some use cases ready it is important to introduce Open RAN now and to evolve it in a pragmatic way to ensure they can coexist with traditional RAN solutions

  • Private 5G Networks
  • Rural Deployment e.g 8T8R FDD
  • In building solutions
  • Macro sites

TIP Open RAN Project Progress

TIP A.K.A Telecom Infra Project is an Open-source project looking after a number of disruptive solutions that can make 5G networks both cost efficient and innovative. Below are some highlights on our progress in the community till 2021

A1: Built the Reference architecture for Multi vendor validations

Through support of operators, vendors and partners built a complete reference architecture to test and validate the complete stack end to end and SI integration

A2: Built the Reference architecture for Multi vendor validations

Worked to define the define the complete RFX requirements for Open RAN and it can be retrieved as below

TIP OpenRAN_OpenRAN Technical Requirements_FINAL

A3: Use cases of Open RAN success

In 2020 and better part of 2021 the community have worked tirelessly to couple Open RAN with some exciting use cases to capitalize the low hanging fruits of Open RAN as follows

  1. “Context Based Dynamic Handover Management for V2X”
  2. “QoE Optimization”
  3. “Flight Path Based Dynamic UAV Resource Allocation”
  4. “Traffic Steering”
  5. “Massive MIMO Optimization”
  6. “Radio Resource Allocation for UAV Applications”
  7. “QoS Based Resource Optimization”

O-RAN+Use+Cases+and+Deployment+Scenarios+Whitepaper+February+2020

A4: Success through RIC (RAN Intelligence controller)

There are two-fold advantages of RIC introduction in Open RAN architectures mainly first for RAN automation for both Managmeent and KPI optimization and secondly bring new and disruptive use cases through xAPPS and data driven operations including

  1. Smart troubleshooting of RAN
  2. RAN parameter optimization using AI
  3. Capacity predicting
  4. New Use cases by exposing API’s to 3rd part developer

A5: RAN configuration standardizations

Benchmarked a number of RAN configurations to be deployed in the field including

  1. Small cells mainly for SME and in building
  2. Low-capacity Macro (Rural)
  3. High-capacity Macro (Highways)
  4. RAN parameter optimization using AI

Highly encourage to join TIP to learn more

https://member.telecominfraproject.com/

A6: Chip advancements brought prime time for Open RAN with 64T64R trial in 2021

Customers realy on 5G massive multiple-input and multiple-output (MIMO) to increase capacity and throughput. With Intel’s newest Intel Xeon processors, Intel Ethernet 800 series adapters and Intel vRAN dedicated accelerators, customers can double massive MIMO throughput in a similar power envelope for a best-in-class 3x100MHz 64T64R vRAN configuration”

Challenges and Solutions

In last two years we have come far to innovate and experiment with a lot of solutions on Open RAN , the list of issues we solved are huge so lets focus on only the top challenges today and how we are solving them today, let’s say to set a 2021 target our target is by the time O-RAN alliance  freezes coming releases Dawn  (June 2021)  and E Release (Dec 2021) we as a  community are able to fix those top issues as below .I apologize to leave down some less complex issues like deploy Open RAN , focus groups key interfaces specifications status specially momentum around 7-2 etc . I feel I ran of time, pages and energy and I fear I will test your patience with a bigger tome

P#1: Ecosystem issues in Open RAN

Based on our field trial we found for 8T8R we can achieve around 40% of cost reductio with Open RAN with the future transport architectures like “Open ROADM” we can build whole RAN network in open manner and achieve great cost and efficiency advantages. However when components are from different vendors the non-technical factors to make it successful is really challenging e.g

  • How to convince all team including RAN guys that what we are doing is right J and we should get rid of the black boxes
  • How to let all partners work together in a team which are historically competitors
  • Making software integration teams

P#2: Radio Issues in Open RAN

The Site swap scenarios in most brown field environments require efficient Antennas and Radios that are

  • 2G + 3G +4G +5G supported
  • Massive Mimo and beam forming support
  • Low cost

This is a fact that till now most of effort has been on the DU/CU part but now we need more attention solving the radios and antennas issues .

Lesson we learnt in 2020-2021 is that everything can not be solved by the software as finally every software need to run over a hardware . An increased focus on Radio , Antennas and COTS hardware is a must to acceerate any software innovation

P#3: Improve RAN KPI

No disruptive solution of Open RAN acceptable unless it can deliver a comparable performance to legacy systems like coverage, speed , KPI .

To make Open RAN main stream DT, tools, RAN benchmarking all need to be addressed and not only the Cloud and automation part 

P#4:SI and certification process

We already witnessed a number of SI forms and capabilities during NFV PSI however for a disruptive solution like Open RAN it need a different approach and SI should possess following

  • Complete vertical stack validation

Its not just the Cloud or the hardware but the end to end working solution that is required

  • Stack should include Radios and Hardware

Certification should consider RF/radio and hardware validation

  • Software capability and automation

To make it successful it is very important that SI is rich on bothy tools and capabilities on automation ,data and AI

Source: mavenir

P#5:Impacts of Telco PaaS and ONAP

To make Open RAN real success it is very important to consider this while building capabilities and specifications of other reference Architectures most of which are Telco PaaS and ONAP . If i go to explain this part i fear the paper length will become too long and may skew towards something not necessarily a RAN issue .

However just to summarize the ONAP community has been working closely with Open RAN to bring the reference architecture in upcoming release

and see some of agreed WI’s in below

https://wiki.onap.org/display/DW/5G+Use+Case+Meeting+notes+for+May+to+Dec+2020

Finally for Telco PaaS we are also working to include Telemetry , Packaging and Test requirements for Open RAN stack . Those who interested in these details kindly do check my early paper below

Open RAN a necessity for 5G era

Early experience with 5G proved that it is about scale and agility, with cost factors driving operators towards an efficient delivery model that is agile, innovative and that can unleash the true potential of network through Data and AI.

In additon as time will pass and more and more use cases will require integration of RAN with 3rd party xAPPs it will be definite need to eolvve to a architecture like Open RAN that will not only support coexistence and integration with legacy systems but also support fast innovation and fleibiltiy over time  . With early successful deployments of Open RAN already happened in APAC and US its improtant for all industry Catch the Momentum

Those who are proponents of closed RAN systems often say that an Open system can never compare with monolithic and ASIC based systems performance, similarly they claim the SI and integration challenge to stitch those systems far outweigh any other advantage.

The recent advantages in Silicon like Intel FLEX architecture with ready libraries like OPENESS and Open Vino has really made it possible to achieve almost same performance like monolithic RAN systems.

Above all latest launch of #intel 3rd generation XEON processors are proving to be a game changer in bringing COTS to the last mile sites .

 Above all involvement of SI in the ecosystem means the industry is approaching phase where all integration can be done through open API and in no time making true vision of level4 autonomous network possible. 

TermDescription
DUDistributed Unit
CUControl Unit
CP & UPControl Plane and User Plane
A & AIActive and Available Inventory
CLAMOControl Loop Automation Management Function
NFVINetwork Function Virtualized Infrastructure
SDNSoftware Defined Networks
VLANVirtual LAN
L2TPLayer2 Tunneling Protocol
SBIService Based Interface
NRFNetwork Repository Function
NEFNetwork Exposure Function
NATNetwork Address translation
LBLoad Balance
HAHigh Availability
PaaSPlatform as a Service
ENIEnhanced Network Intelligence
ZSMZero touch Service Management
EFKElastic search, Fluent and Kibana
APIApplication Programming Interface

5G Cloud Edge Services and MEC Open RAN Orchestration RAN

Saad Sheikh View All →

I am a Senior Architect with a passion to architect and deliver solutions addressing business adoption of the Cloud and Automation/Orchestration covering both Telco and IT Applications industry.

My work in carrier Digital transformation involve Architecting and deploying Platforms for both Telco and IT Applications including Clouds both Open stack and container platforms, carrier grade NFV ,SDN and Infra Networking , DevOps CI/CD , Orchestration both NFVO and E2E SO , Edge and 5G platforms for both Consumer and Enterprise business. On DevOps side i am deeply interested in TaaS platforms and journey towards unified clouds including transition strategy for successful migration to the Cloud

Please write to me on snasrullah@swedtel.com

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: