Defined by the Third Generation Partnership Project (3GPP) at the beginning of this century, SON (self-organizing network) algorithms configure and optimize mobile networks automatically without human interaction. As networks handle vast amounts of traffic and become more and more complex through the addition of several generations, SON algorithms are blended with machine learning (ML) and artificial intelligence (AI) algorithms. This report tracks the centralized SON (C-SON) software and services revenue market, provides size, forecast and vendor market shares, and analyzes the evolution of C-SON modules and use cases becoming apps in the RAN Intelligent Controller (RIC) needed in open RAN, open virtual RAN, and software defined RAN (SD-RAN) architectures. This report also tracks the O-RAN defined service management and orchestration (SMO) market and provides a breakdown between non-Real Time RIC rApps and near-Real Time RIC xApps.
TABLE OF CONTENTS
ABSTRACT
KEY TAKEAWAYS
IT'S AMAZING, THE AI-POWERED SON KEEPS SHINING!
- DEFINED BY THE NGMN IN 2008 AND STANDARDIZED BY THE 3GPP IN REL.8, SON HAS BEEN AROUND FOR OVER A DECADE AND TOOK OFF WHEN LTE ROLLOUTS STARTED IN 2010
- Already in 2010, zero touch network closed-loop automation was the C-SON grand vision
- IN 2023, GLOBAL SON SOFTWARE AND SERVICES REVENUE GREW 2% YOY
- This year, we forecast the global SON market to keep growing at low single digits
- In 2024, EMEA and Asia Pacific will more than offset the decline in North America
- C-SON sales have never grown proportionally to the number of mobile networks
- With or without C-SON, LTE networks are running fine, so are 5G NSA ones
- If every single LTE network was equipped with a C-SON, we believe global C-SON software and services would be more than $8B at this point
- This 2024 peak coincides with the open vRAN uptake
- C-SON has always been a tough sell
- As a result, the C-SON ecosystem has shrunk to 10 active players
- While focusing on Non-RT RIC, most players are still investing in SON use cases
- Our estimates indicate Nokia continues to lead C-SON sales market shares
- WITH THE EMERGENCE OF OPEN RAN, SON USE CASES ARE MORPHING INTO APPS
- Since 2017, AT&T has been driving the ONAP and open RAN efforts
- Fueled by geopolitics, open RAN accelerated the development of the RIC and the SMO
- Leveraging the TIP and the O-RAN Alliance, the ONF SDN'ized the RAN
- Deutsche Telekom rushed to trial SD-RAN but that was it!
- In April 2022, Intel acquired the ONF development team and Ananki
- in 2021, we developed a RIC and SMO model based on the O-RAN Alliance framework
- In August 2023, the O-RAN Alliance released a vision of the NIC, the evolution of the RIC
- In the meantime, open RAN networks run fine with C-SON, which is here to stay longer than many think
- Only Dish operates a 5G-only open RAN footprint, open RAN started in 4G LTE
THE LULL IS OVER, AT&T IS CHANGING THE GAME, THE NON-RT RIC, ITS RAPPS AND THE SMO ARE TAKING OFF
- IT'S UNBELIEVABLE, THE C-SON MARKET DEVELOPMENT HISTORY REPEATS ITSELF
- BAKED IN 5G OPEN RAN, RIC IS COMPLEX AND IS HEADING FOR PRIME TIME, THANKS TO AT&T FOR TAKING THE PLUNGE
- Like its C-SON predecessor, RIC uses AI and ML, and needs to send back instructions to the RAN
- Consequently, the Non-RT RIC is taking off for good
- In 2027, we expect 100% of C-SON sales and services to be in the form of a non-RT RIC rApps
- The xApp will have to wait because the training requires a vast amount of RAN scenarios that can be generated to reproduce real events
- The Near-Real Time RIC market is unleashing D-SON use cases that until now were proprietary and controlled by the RAN vendors; they still are though!
- CSPs have plenty of questions about the near-real time RIC
- The App ecosystem is mushrooming...
- ...some, particularly in the RIC domain, won't make it
- ALL THIS DRIVES THE SMO MARKET
- Coming out of a lull, the SMO market is poised for a rapid ramp up, and will be the largest opportunity as it gradually replaces the OSS
- Asia Pacific will be the largest SMO market throughout the forecast period