Is OpenRAN (Open Radio Access Networks) the path to saving costs in radio access networks? Vodafone seems to think it is as it trials the technology in its networks in the UK and Africa. Among Vodafone’s goals are: 1. Improve supply chain resilience; and 2. Connect rural communities using standardized, lower cost network equipment. This is good news for OpenRAN ecosystem companies such as Mavenir, Parallel Wireless and others. But I think Nokia, Ericsson and Huawei are not panicking as Vodafone drives to empower other vendors. This is because there are a few serious obstacles in the path of OpenRAN to commercial deployments.
Shrinking Number of Players
The number of RAN players has shrunk to a handful: the top three – Huawei, Ericsson and Nokia – control over two thirds of market share. With the US pushing to isolate and breakdown Huawei, operators, especially those in US-allied countries, are facing tough choices on who to chose for their network evolution. Vodafone is facing similar tough choices as it plans to evolve its global network in markets with very different economics. Isolating Huawei not only hurts it, but it adversely impacts operators worldwide as it leaves them with dangerously few options. Already operators in markets such as Canada have implemented purchasing freezes as did many others in Western countries.
Based on the above, the problem in radio access networks is the limited options available to operators. This tilts the balance of power in favor of the vendors. The actual capex operators pay for RAN is only 10-15% of the total cost of network ownership. It is lower performance and higher opex that operators will suffer most from.
The Challenge with OpenRAN
The idea of OpenRAN is very attractive. Decomposing the radio network into multiple nodes connected by well-defined interfaces and running on COTS hardware is the antithesis to today’s RAN which is based on vendor-proprietary hardware and optimized non-interoperable interfaces [see here and here for background]. Theoretically, opening the radio access to multiple competing vendors should lead to lower cost. But in reality this is not always true: there are situations where an ‘open’ infrastructure results in higher cost and lower performance. Consider these two reasons:
- The Test and Verification Challenge. One challenge that is rarely recognized is the extent of testing infrastructure products require before being deployed in the field. This is especially the case in radio access products where tolerances are much smaller than in the core network. A ‘best-of-breed’ RAN has never existed! To realize one would require vendors to engage in extensive testing with limited ability to control a partner’s roadmap. The concept of ‘plug-and-play’ RAN is very difficult to achieve. Consider this challenge from both the operator and the vendor perspectives to glimpse the complexity. Both would be facing challenges in realizing such a network.
- The Responsibility Challenge. A single RAN vendor presents an easy case for the operator on support and maintenance. When a failure occurs, the operator asks the vendor to fix the problem per the service SLA. But in OpenRAN, the situation is more complex as it depends on the system integrator and suppliers of different network elements. How will operators structure the SLAs and related liabilities? Who will take responsibility and stop the finger-pointing that will result when things go seriously wrong?
These are only two of several reasons that contribute to higher costs. Huawei for its part has been vocal that virtualized and multi-vendor sacrifices efficiency and performance. Other vendors share the same sentiment but have not been vocal.
OpenRAN Success Factor
Overcoming OpenRAN challenges is possible, but requires strong and active operator engagement without which OpenRAN will be doomed to failure. Operator engagement – being active in the driver seat – is critical. I believe to make the market happen the operator needs to assume leadership role in system integration. This is a major change for the operators who spent years squeezing vendors in contract negotiations. Operators did this in part because they themselves lacked the ability for system integration, and many even outsourced network support and maintenance to their vendors under managed services contracts. This will involve a cultural change before anything else.
Rakuten is the first to implement a multi-vendor RAN, using products from Altiostar, Nokia, Quanta and others. They are the company to watch even as their program faces challenges and delays. Rakuten had to delay the launch date from this month, October, to April as they struggle with network roll-out. Verizon would be another operator to watch.
OpenRAN, similar to network virtualization, will only succeed if operators muster the stamina to go through a period of pain. In my opinion, this makes OpenRAN expensive in the short term – not for the leading vendors to be concerned about. But looking at the long term, OpenRAN has its benefits. So, are the operators up to the challenge?