Tag Archives: RAN

A Look at Open Ran

AT&T recently announced a $14 billion dollar deal with Ericsson to start upgrading its cellular networks to Open RAN (ORAN). I suspect most of the readers of this blog know a lot more about fiber technology than cellular technology. They’ve probably been seeing headlines for years talking about various flavors of RAN technology and probably wondered what it means. The bottom line is that the AT&T announcement is a big deal and means a big shift in the cellular vendor industry.

RAN stands for Radio Access Network, which is an acronym for the radios and antennas at a cell site. Open RAN represents a migration to a software-driven network where any brand of radio can be used at a cell site. This is a big deal in an industry where three major companies – Huawei, Ericsson, and Nokia – have made the gear for cellular networks. This means that most cellular networks rely entirely on the proprietary gear of a single vendor. A shift to Open Ran is akin to the migration a few years ago when data centers migrated to lower-cost white box routers and switches instead of having to buy proprietary gear from large vendors like Cisco.

However, the migration of the cellular network to generic hardware is a lot more complicated than making switches operate in a data center. This is mostly due to the complicated nature of operating a cell site in an environment where factors like temperature, precipitation, and customer traffic volume change the overall performance and operating characteristics of a cell site during the day. I always think of the old saying that operating a fiber network is science while operating a wireless network is art. The only solution for dealing with the complexities found at cell sites has been to pick an integrated suite of products from one vendor.

AT&T is the largest cellular carrier to announce it is moving towards ORAN. A few years ago, the newly formed DISH cellular effort announced it would use ORAN, but the technology was not ready, which drastically hindered the company’s initial launch of the network. The UK announced in 2021 that it has a goal to migrate its cellular networks to 35% Open RAN by 2030. The AT&T announcement doesn’t mean the company will be able to freely integrate any brand of radios into its networks but is a big first step towards true Open RAN.

The original view of Open RAN was that a cellular company could put any radio anywhere, connect it to the backhaul, and it would work. The software in the cloud would automatically handle the integration with the overall network and the surrounding cell sites, and the new site would work out of the box.

We’re still a long way from that ideal and may never fully get there. Cellular carriers have relied on the major vendors because they made sure that a cell site would work. I think back to all of the other launches of wireless technology where vendors released beta equipment and the first generation of customers were, unfortunately, the guinea pigs. I painfully recall watching ISPs try to launch wireless broadband networks using LMDS and MMDS spectrum and failing when the radios didn’t operate as promised.

Part of the reason that the migration to Open Ran has been slow is that vendors and carriers have also been pursuing other network options to cut costs. Historically, all of the electronics were fully integrated at a cell site. The RAN radio units were installed on a tower, and the BaseBand unit (BBU) with all of the brains was installed in a hut at the base of each tower. Technicians had to go through the time-consuming process of fine-tuning the electronics at each tower to meet the specific local circumstances.

In recent years, there have been efforts to consolidate some of the BaseBand Unit functions.

  • DRAN (Distributed RAN) was the first attempt to separate the base electronics from the radios. That was a lot harder than it might seem because a cellular tower relies on precise timing and fiddling with the type or length of wiring between the radios and the base caused all sort of problems.
  • CRAN (concentrated RAN) moved the base units into regional data centers (BBU hotels) and opened up the concept of DAS – Distributed Antenna Systems that would work from these centralized sites.
  • C-RAN (Cloud RAN), not to be confused with CRAN, has been a movement to move the entire brains of a cell site into data centers.

Open RAN takes something from each of these various evolutions and takes the final step of breaking the equipment monopoly of the big vendors. This isn’t going to be easy and has the danger of moving back to a time when we test new radios by foisting them on the public. But the savings from ORAN are potentially gigantic, and if we’ve seen anything in the wireless industry, it’s that the lowest-cost option is going to get the most attention.

More on 5G Standards

I wrote a blog last week about the new 5G standard being developed by the International Telecommunications Union (ITU). This standard is expected to be passed this November. However this standard is not the end of the standards process, but rather the beginning. The ITU IMT-2020 standard defines the large targets that define a fully developed 5G product. Basically it’s the wish list and a fully-compliant 5G product will meet the full standard.

But within 5G there are already a number of specific use cases for 5G that are being developed. The most immediate three are enBB (enhanced mobile broadband, or better functioning cellphones), URLLC (ultra-low latency communications to enhance data connectivity) and mMTC (massive machine type communications, to communicate with hordes of IoT devices). Each use case requires a unique set of standards to define how those parts of the 5G network will operate. And there will be other use cases.

The primary body working on these underlying standards is the 3GPP (3rd Generation Partnership Project). This group brings together seven other standards bodies – ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC – which demonstrates how complicated it is to develop a new wireless technology that will be accepted worldwide. I could talk about what each group does, but that would take a whole blog. Each standards group looks at specific aspects of radio communications such as the modulating schemes to be used, or the format of information to be passed so that devices can talk to each other. But the involvement of this many different standards groups explains a bit about why it takes so long to go from a new technology concept like 5G to functioning wireless products.

There is currently a lot work being done to create the specific standards for different portions of a 5G network. This includes the Radio Access Network (RAN), Services and System Aspects (SA) and Core Network and Terminals (CT).

The 5G RAN group, which looks at radio architecture, began work in 2015. Their first phase of work (referred to as Release 15) is looking at both the eMBB and the URLCC use cases. The goal is to define the specific architecture and feature set that is needed to meet the 5G specification. This first phase is expected to be finished in the fourth quarter of 2018. The 5G RAN group is also working on Release 16, which looks more specifically at getting radios that can comply with all of the aspects of IMT-2020 and is targeted to be completed in December of 2019.

The 5G SA group has already been actively working on the services and systems aspects of 5G. The preliminary work from this group was finished last year and final approval of their phase 1 work was just approved at the Mobile World Congress. But the SA group and the RAN group worked independently and it’s expected that there will be work to be done at the end of each phase of the RAN group to bring the two groups into sync.

The work on the core network has begun with some preliminary testing and concepts, but most of their work can’t be started until the RAN group finishes its work in 2018 and 2019.

The reason I am writing about this is to demonstrate the roadblocks that still remain to rolling out any actual 5G products. Manufacturers will not commit to making any mass-produced hardware until they are sure it’s going to be compatible with all parts of the 5G network. And it doesn’t look like any real work can be done in that area until about 2020.

Meanwhile there is a lot of talk from AT&T, Verizon and numerous vendors about 5G trials, and these press releases always make it sound like 5G products will quickly follow these trials. But for the most part these trials are breadboard tests of some of the concepts of the 5G architecture. These tests provide valuable feedback on problems developed in the field and on what works and doesn’t work.

And these companies are also making 5G claims about some technologies that aren’t really 5G yet. Most of the press releases these days are talking about point-to-point or point-to-multipoint radios using millimeter wave frequencies. But in many cases these technologies have been around for a number of years and the ‘tests’ are attempts to use some of the 5G concepts to goose more bandwidth out of existing technology.

And that’s not a bad thing. AT&T, Verizon, Google and Starry, among others, are looking for ways to use high-bandwidth wireless technologies in the last mile. But as you can see by the progress of the standards groups defining 5G, the radios we see in the next few years are not going to be 5G radios, no matter what the marketing departments of those companies call them.