TECH TALK: Is Open RAN Really the 5G Panacea?

By Ernest Worthman, Executive Editor, AWT magazine, Senior Member, IEEE

November 24, 2020

Worthman

Open Radio Access Networks (RAN) are being touted as one of the most promising technologies for accelerating and simplifying 5G deployments. Ajit Pai, the FCC’s Chairman, claims it will transform 5G network architecture, costs and security. The China Hawks on the Hill want to throw $750 million for ORAN R&D at it (which has since been reduced to $50 million) and, of course, has been justified by the usual anti-Chinese security rhetoric. However, the government throwing money at it does not necessarily mean it is solid and there are no risk factors. And is open-source the golden ring of hardware interoperability?

There has been much activity in this segment and some, such as Pai, have touted it as the ideal platform for 5G, going forward. But it is not being fast-tracked by any means. There is, in fact, plenty of controversy around it.

Support for it abounds. If one looks at the ORAN Alliance, it reads like a who’s who of telecom. There is no shortage of players (230+ globally) and the hype coming out of it is certainly powerful.

The arguments for ORAN are certainly compelling. A hardware platform that allows for vendor-neutral equipment to be integrated into any deployment. The theory is that all one has to do is design the network with ORAN standards and the hardware can come from any vendor that builds to the standard. Sounds awesome! But such concepts are far from new and no matter what industry it is, not everybody goes running out and jumps on the ship.

This is a familiar scenario played over and over. If we go back to the birth of modern day cellular, circa 1984, the world had a chance to create a one-platform cellular world. However, it did not go that way. Self-interests ended up creating a two technology platforms world – CDMA and GSM.

Next, take a look at the mobile operating systems (MOS). At last count there are 15 different systems available. Similar arguments can be made for the PC industry – not just PCs but OS’s and hardware platforms, as well.

If one follows the segment, one sees a bit of disquiet, regardless of what Pai thinks and promotes, or how many vendors are connected to the platform. There are certainly those willing to go with it, but others are seeing the glitches.

For example, a couple of months ago Vodafone turned on its first Open RAN site at the Royal Welsh Showground in Powys, Wales. As well, another U.K. MNO, Pilot, will be scaling up to as many as 2,600 Open RAN sites in Wales and England. Other projects are going on, as well, around the globe.

ORAN, or disaggregated RAN, does offer a number of compelling advantages as does any open platform. They all relate to the fact that it should be simpler to deploy and cheaper due to the vastly expanded vendor ecosystem.

In general, there is collective agreement that ORAN is a good way to proceed with 5G deployment. Theoretically, it should simply be a matter of plugging in the hardware one desires. However, as I like to say, all that glitters is not necessarily gold.

One of the more visible issues is system integration. Being able to simply pick off the shelf products from any number of vendors and being assured the product will plug and play, is certainly attractive. But that has to be a given. Unless 100 percent system integrability is assured there will be resistance. And, so far, system integration has some challenges.

First of all, while 5G is the primary target for ORAN, the current wireless infrastructure is still a bit of a hodgepodge with both 3G and 4G, and now, nascent 5G technologies. The sun is setting on 3G, but it will be a while before it disappears. Thusly, ORAN systems need to be able to handle them all, at least for a time.

Before ORAN, proprietary systems were designed for a specific purpose with the best hardware and software to do the job. The MNOs built their systems with parts from particular vendors that have been certified to work together. There were no integration issues since the RAN was a closed package. The baseband unit, for example, had the radio head and all the layered hardware to put the data out to the evolved packet core.

With ORAN, that becomes a sliced-out structure with individual components, the radio unit (RU), distributed unit (DU), and the centralized unit (CU). There are also open interfaces between the RU, DU, CU, radio access point (RAP), RAN, and the radio intelligent controller (RIC). And do not forget the role virtualization is expected to play. There is software in the mix, as well.

Some of the challenges are minor, such as the physical interfaces between the components or layers. That will require a robust test and certification process, which has yet to be firmly established.

However, the biggest challenge with ORAN is managing open interfaces from multiple vendors. Among them are security, design specifications, ownership (not physical but failure accountability), reliability, vendor support, specific point of failure, and more. This all adds up to additional layers of complexity, particularly with a diverse maze of vendors.

At the top of that pile is security. Overall security concerns simply ratchet up just because of the increase in components. Closed RAN simply meant one had to secure the input and output ports. ORAN creates a slew of new security portals – APIs, real or near real-time RICs, supply chain validation, open-source code, even hackers.

Will Townsend, one of the analysts I occasionally cross paths with notes that “industry-standard compute, storage and networking elements are thrown into the mix, thereby increasing the overall threat surface.” I agree with his perspective on this.

However, despite the multitude of challenges, ORAN is moving forward and seeing early systems based on early components. Looking back from the end game once all the challenges are mitigated, it does promise to bring disruptions. On top of that list is economic, which is always the primary driver. Early models show as much as a 30 percent savings in CAPEX.

Currently, it is especially attractive to greenfield operators such as Dish here in the United States and operators in other countries where the infrastructure is not quite as developed. But there are still unknowns around the cost of integration and ongoing maintenance once the system is deployed.

In the end, ORAN makes sense. However, dreams of open networks, platforms and systems have all been envisioned in the past. Success has been varied. Not all have proven to be a panacea for proprietary systems. In the telecom space, open systems are an elegant solution if everybody is going to use them. The less the network costs to operate, the lower the costs to the consumer (if you believe that is how it will go, I have some swampland in Nevada I will sell you).

Seriously, that is usually how it ends up, to some degree. But we all know that MNOs have been somewhat surreptitious in paying cost savings forward and being above board with charges. The only thing that will benefit the consumer is competition. ORAN does promise to make deployments quicker and cheaper so hopefully, that will add competition and some of that will roll downhill.

Well, we will be dark on Thursday, so here’s wishing all of you a very happy Thanksgiving!