We connected an island.
An actual island. Surrounded by water and everything. Networking people use “island” as a metaphor so often that occasionally we have to clarify.
One of our service provider customers needed to connect their Ethernet network on an island to the US. They used our Path Border Controller, the PBC 2000, to extend their private network over already available transport, without building new fiber.
It worked. The customer could see the value.
Then came the question every infrastructure company should want to hear:
What else can we sell over this?
That question led us to the PBC 100 Series.
And the answer turned out to be a lot.
The Challenge: The Last Mile Decides What Service Providers Can Sell
The provider wanted to deliver more to its customers. But the last stretch, from each customer’s premises back to the provider’s point of presence (PoP), made that harder than it should have been.
Every new service meant reconfiguring the network between the customer and the provider. Services that rely on multicast meant PIM and the routing configuration around it. Others meant another round of NAT and WAN changes. And keeping each customer’s traffic separate added more work on top.
Customers wanted another service. The network wanted another change window.
This wasn’t just an island problem. Every operator has a map on the wall: lit buildings in one color, everything else in a polite shade of “not yet.” And even inside the lit part, the last stretch to the customer quietly decides what you can sell.
We’d made it easier to connect the operators’ networks. Now we needed to make it easier to deliver services all the way to the customer.
The Goal: Bring the Customer’s Ethernet to the Services
What if the operator could treat a remote customer site as an extension of the Ethernet network at its PoP?
Keep each customer’s traffic in its own segment. Connect that segment to the service it needs. Deliver the service from the operator’s own network or a connected partner, without redesigning the customer’s WAN connection every time.
The customer premises doesn’t have to become a small branch office of every service vendor.
That changes the operator’s job. Instead of squeezing another service through a pile of routing and NAT rules, it connects the customer’s Ethernet segment to the right service.
The Ethernet connection doesn’t eliminate the work of configuring a service. It gives that work a consistent place to happen.
The Solution: The PBC 100 Series
To extend Ethernet directly to the customer’s site, we needed another Path Border Controller at the customer premises. The PBC 2000 already handled the operator side, but customer sites are a different setting. Some have modest bandwidth. Some have no fiber at all. Putting a PBC 2000 everywhere wasn’t the answer.
So we built the PBC 100 Series.
Small Device, Bigger Map
- PBC 100 Series at the customer premises.
- PBC 2000 at the operator’s PoP, Metro Ethernet edge, or data center.
- In between: available IP transport, whether fiber, dedicated internet access, satellite, or a combination.
The customer site just needs working IP connectivity. Our cloud-based Path Computation Engine computes the path, and the Path Border Controllers establish the encrypted Layer 2 connection and measure the service.
The customer gets an Ethernet handoff. The operator gets that customer’s segment extended into its network, without a new fiber build or a manually negotiated NNI for that connection.
The map still matters. It just gets less say in the sales meeting.
Why Ethernet Matters Here
The provider needed more than a path between sites. It needed a way to connect each customer to services delivered from its network.
That’s what SD-Ethernet does. It extends the customer’s Ethernet segment to that service delivery point, keeping customer segments separate and giving the operator a consistent handoff to build on.
For multicast, the customer-to-PoP stretch carries Ethernet frames without requiring a separately routed multicast design. For new services, the operator can connect the customer’s segment to the appropriate service, reducing repeated NAT and WAN changes at the premises.
That means less engineering time per service, faster turn-up, and the ability to serve off-net customer sites.
The transport provides reach. The Ethernet service provides the handoff the operator can build on, turning reach into revenue.
How It Works: The First Connection Opens the Next Opportunity
Here’s how it plays out.
- Connect the site. A PBC 100 Series device at the customer premises pairs with a PBC 2000 at your PoP, over whatever IP transport the site has. The site doesn’t need to sit inside your fiber footprint.
- Land the customer’s Ethernet at the PoP. That customer now has its own Ethernet segment, extended into your network and kept separate from everyone else’s.
- Attach services to that segment. Cloud access, private connectivity to partner networks, access to AI infrastructure, from your own network or MaiaEdge-enabled partners. Services can live at the PoP or be reached through it.
- Add the next service from the PoP. Use the existing Ethernet segment to connect the customer to another service, reducing the need for repeated WAN and NAT changes at the premises.
Once that first PBC 100 Series to PBC 2000 connection is up, you have a consistent way to deliver more services to the site.
The destinations keep moving: a GPU provider today, a different data environment tomorrow. Each becomes another service opportunity over the connection you’ve already built.
The first connection gets you to the customer. What you deliver next grows the business.
Available Now
The PBC 100 Series gives operators a way to grow their addressable market without expanding their physical network, and to deliver more services to the customers they reach.
The first device in the series, the PBC 124-C, is available now. Explore the complete Path Border Controller family, or bring us a customer site you’ve struggled to serve.
We connected the island. The next step was bringing more services ashore.
Book a Meeting
