Dish's Marc Rouanne: Building an open RAN, greenfield 5G network

Dish Network has been slowly assembling the key vendors it plans to use to build and launch a nationwide 5G network in the US. The company is using open RAN technology to build a new, "greenfield" network from scratch with its massive lowband and midband spectrum holdings.

This week the company announced Altiostar would join Mavenir in managing its radio access network (RAN), and that Japan's Fujitsu would provide the physical radios that will broadcast Dish's 5G signals.

Marc Rouanne was hired late last year to construct Dish's 5G network, and has been working out of Dish's headquarters in Denver with the rest of the company's network team to establish the basic outlines of Dish's network design.

This discussion with Rouanne has been edited for clarity and brevity.

Light Reading: It appears that Altiostar and Mavenir are playing basically the same role in Dish's network design. Is that right?

Rouanne: There are some functions where we may have vendors for the same functions, and we will consume based on the geography – but it also could be the [network] slices – over time. So there's a unique opportunity for us compared to the existing architecture.

At the beginning, it will be similar [when using each vendor]. But I expect that, over time, I would not be surprised that we start seeing some more specialization [between the two vendors] and more capabilities.

It's a bit like you would use different antennas and radios. When you mix them, all of a sudden you have your network benefiting from the best.

We're not going to use the same capabilities [from each vendor] for each slice, over time.

Light Reading: What can you say about the radios that Fujitsu will provide?

Rouanne: These are the best radios we've found. And we're delighted with the technology.

This is classic 5G. It's not like we're doing millimeter wave [which is the spectrum that Verizon and others are using for 5G]. When it comes to those spectrum bands [owned by Dish] and open RAN, this is clearly the best-in-class we've seen. And we looked at dozens of vendors. Many, many vendors.

Light Reading: Could Dish add additional radio vendors to its lineup?

Rouanne: Yes. Anyone can enter at any point. We'll continue to look for advanced radio from anyone, and we're talking to everyone in the market.

Light Reading: Is Fujitsu capable of supplying all the radios Dish will ultimately need?

Rouanne: I'm not worried.

Light Reading: There have been some questions about whether open RAN technology – including the "option 7" version that Dish and Japan's Rakuten are using – is ready for prime time. What would you say to that?

Rouanne: When you ask that question, I think we have to be specific. Can all the vendors apply it? When people say, "No, we cannot do open RAN," you have to ask, "When and why? What spectrum, what radios?"

For the types of solutions we use, it's available today. And we see the benefits. I mean, we are getting what we want, what we need. The radio is pretty cool.

Light Reading: What should Dish's 5G customers expect in terms of network performance and speeds?

Rouanne: What we are generating is a network with different SLAs [service level agreements]. So it's a bit funny that you're asking me this because that's not the way we think of it. Because we'll have different SLAs. And in the different SLAs, we'll have different levels of service capabilities.

We think that when we define these, it's going to change the way people measure networks. Because if you measure the SLA of a healthcare or automotive [network] slice, would you measure it in the same way as streaming [video] to a consumer? I guess not.

We're designing the network so that we can offer at the right locations the SLAs that will be required for different vertical industries.

Light Reading: Will Dish continue to announce other vendors for its network? And what other types of vendors is it looking for?

Rouanne: We're working with many, many vendors.

We made two architecture decisions: One was to go with open RAN, and you're starting to see the results of that. The second one was to go with a cloud-native-capability network, where we put on top of our cloud CNFs [containerized network functions]. So the core vendors are CNFs. And what we've seen is that the maturity of the market is remarkable.

We have multiple choices, many choices of containerized network functions, and we've been testing. Then it's going to be a matter of deciding when we need to make a decision, who do we embark?

But those two decisions we've made have proven to be pretty solid. I think we are happy with the progress we've made on testing the market.

Light Reading: What are Dish's next steps for 5G in 2020?

Rouanne: We wanted first to have the brain of the radio – which is the Mavenir, Altiostar – which helped us define the specification. And then we needed to have the equipment; that's Fujitsu. So we are just building the things as we need to consume them.

We're taking the time we need.

Light Reading: When might Dish first launch commercial 5G service?

Rouanne: We're not sharing that.

— Mike Dano, Editorial Director, 5G & Mobile Strategies, Light Reading | @mikeddano