By late 2026, roughly 20,600 private 4G and 5G networks will be in service, according to Analysys Mason. That’s nearly 2,500% more than at the end of 2020. Impressive? Yes. Surprising? No. In a Capgemini Research Institute survey three years ago, enterprises were already champing at the bit to begin using private 5G.
Enterprises have three main options for implementing a private 5G network:
- Get a spectrum license, such as CBRS in the U.S., and then buy and operate a RAN and core. This gives the enterprise complete autonomy —if they’re willing and able to climb the steep learning curve that comes with owning and operating a mobile network, which is fundamentally different from, and far more complex than, the Wi-Fi networks many have.
- To avoid that cost and complexity, many enterprises will choose the second option: Using a private slice of a public 5G network.
- Some enterprises will go with the third option, where they own a network but also use a private slice of a public network. An example is a logistics company that owns the network at its warehouses and other facilities but also uses a private slice of the public network to provide service away from those facilities.
Neutral host providers enter the market
But there’s actually a fourth option for implementing private 5G: neutral host providers, similar to the ones that have been providing 3G, 4G and Wi-Fi networks in public places such as malls and airports. To understand where they fit into the private 5G market, consider the example of a university, where faculty, staff and students are customers of three operators.
Those operators obviously will want those customers to have solid coverage and ample bandwidth when they’re on campus. But if all three extended their 5G RANs to that university, it could easily add up to hundreds of cell sites on campus. In fact, if it’s a large, densely populated campus,one operator alone might have 300-plus sites due to 5G’s small cell architecture. That would push the grand total for all three operators to around 1,000.
That model simply doesn’t make sense for the university or for the operators. For example:
- It’s tough to hide all of those antennas, radios and cabling so they don’t undermine the building architecture that universities prize.
- It’s expensive for each operator to buy, install and maintain all of that equipment - capex and opex that makes it difficult for them to sell their services at rates that are competitive yet profitable.
- A thousand sites shoehorned into, say, a 300-acre campus makes it likely that many of them will interfere with one another. Tracking down and mitigating that RFI takes time and money.
- Some faculty, staff and students might raise concerns about potential adverse health effects due to such a high density of sites - concerns that are already amplified with 5G.
A neutral host provider mitigates all of those challenges. A single set of infrastructure and frequencies serve all customers of all three operators. Another potential benefit is coverage in university-owned arenas and stadiums, where mobile operators are not permitted to deploy their sites. Neutral host providers don’t have that limitation, which means they’re a way for operators to serve their customers at games and concerts.
The role of hyperscalers
Hyperscalers will play a key role in this model by providing the cloud infrastructure that neutral host providers will use to support each operator customer. For example, instead of setting up a network operations center (NOC) in each location, neutral host providers will use the cloud to provide that to multiple operators in multiple markets.
Another key role is local offloading in local private clouds. For example, if a large hyperscaler has a cloud network, it can always set up a local, private mini cloud inside, say, a university. And if the university’s neutral host provider wants to set up any 5G-powered systems, that local cloud will help. The hyperscaler will take care of connecting the local cloud to the public cloud, freeing the neutral host provider from that task.
But there are a few caveats and considerations. For example, hyperscalers typically have no experience with telecom-specific requirements that affect a neutral host provider’s ability to meet its SLAs with mobile operators. Some cloud providers automatically reboot their servers every eight hours or so, a practice that’s directly at odds with the telecom standard of 99.999 percent uptime.
As a result, neutral host providers must educate their hyperscaler partners about these and other requirements, such as latency and security. For example, hardware accelerators may be required to support latency-sensitive customer applications such as multi-player gaming or industrial robotics.
The bottom line is that there are multiple options for meeting the burgeoning demand for private 5G networks. By partnering with hyperscalers, neutral host providers can extend their public 3G, 4G and Wi-Fi business models to private 5G.
