Understanding NSN’s performance quality strategy

Does fifty product announcements bring clarity or risk obscuring the message? Keith Dyer talks to NSN's VP of Portfolio Management Thorsten Robrecht.

It is Tuesday afternoon at Mobile World Congress. Your writer is wondering aloud exactly how many announcements Nokia Solutions & Networks (NSN) has made before and during the event. It seems the company has been firing off product press releases for three weeks.

NSN’s VP Portfolio Management, Thorsten Robrecht, chuckles and says the vendor has had 50 – he spells it out – “Five-Zero” product launches around MWC. Well, TMN has got just shy of half an hour for this briefing, so there’s no point trying to list them. Instead, we ask Robrecht to pick up on some key areas and explain what NSN is doing differently in each area compared to other vendors. Overall, if you want to understand NSN’s key top line message, it’s about performance quality.

Too long didn’t read version:
1. We may be later than some to small cells, but ours are built with the exact same software as our macro cells, which is a big operating advantage.
2. Our liquid apps server is being demoed and white labelled by Vodafone. That’s great validation. Combine our small cells controller and liquid apps and you’ve got something very different.
3. The market isn’t “getting” active antennas quite yet but we’ve taken the decision to move to a second generation product in any case. Operations and network planning departments don’t get it yet.
4. We virtualised everything in the core onto genuine COTS. Nobody else has done this. It gives operators an uncrashable core.
5. Network security is getting airtime. We’re now starting a new service in the network to stop malware getting to devices, and then if malware does get through we can block bad content or messages on the uplink from the device.
6. Applied Big Data: we’ve got Minority Report tech for the network, predicting service degradation and network outages 48 hours before it happens, with 95% success rate.

Small cells:
The key differentiator is about having macro feature parity, we’ve been beaten up for being a little bit late with small cells on the market. And we said OK the market is not there maybe, but if we are coming out we are coming out with a clear strategy that is this: “There is no product in this company on the radio access which uses different software or a different chipset, it’s all one glue, one software running in all network elements. Full stop.” That is a big differentiator, no other company is doing this – they’re all making small acquisitions, they bought this company, they bought that company, that product and that product, we are the only one throughout from very big macros down to the small picos indoor [that are] all running exactly the same software. Which means, if you build a network you’ve got one NMS, one feature set, one upgrade, exactly the same services at all places of the network, and higher performance because we are running interference cancellation and so on with all the same algorithms.

(COUNTER point if you shrink down a macro you don’t have something optimised for small cell deployment. The reason there are dedicated standalone small cells is that companies have invested time and money into SON, management, into SoCs optimised for small cell use cases.)

What we did already – five years ago we defined the chipset and software that way that it really scales down, it means it’s not exactly the same chip but a chip family which we defined so that it scales down to power consumption, memory etc So it is not exactly the same part number, but it’s still the same software on all of them.

Liquid Apps:
(You announced LiquidApps and the Radio Application Server a year ago, are you getting anywhere with that?)

We announced KT [Korea Telecom] last year, we finalised the lab activities with them, and are in field trials. Then you might see our exact same product on the Vodafone booth, you see our product there. Vodafone is promoting it themselves, branded it themselves, and we’re very happy they are creating an ecosystem. On top of this, because we are ahead of the pack, we are providing services with IBM on app support, providing video surveillance and caching on two networks. The first app that seemed to resonate reasonably well is caching. Most important thing there is latency improvement. Take the football game experience where you have a stadium with a lot of people and you want to provide in-game video from all angles, it’s there that it’s wanted so create and upload the content there. It saves backhaul traffic, but it’s also a very quick user experience.

So we are progressing, it could be quicker but still it’s a major change. I’m happy that Vodafone is on top of it and promoting it themselves.

We’ve combined this with small cells: that is another differentiator. We have got a central controller point that can control 200 small cells together so they behave like one BTS. You can start some operating independently, and then when you have a lot of small cells you can dynamically re-route IP addresses to this controller remotely and then it all behaves like one BTS. And on top of that in this controller we have this liquid apps server so the same hardware piece that provides advances for small cells can no spread a local content and services capability.

FlexiZone Small Cell Demo

Active Antennas
This is my baby, but this industry is slow sometimes, so we have not so much volume deployments but a lot of small field deployments and trials. We are in second generation development which for me was a crunch – “Was it worth to do it?” I think it’s too early and operators are not picking it up because there’s certain complexity to network planning.

At a high management level the lifecycle benefits are well understood, but go down in the organisation and the network planning tools, structure, thinking is very traditional. People would rather buy something they know. I want to say, “Guys, instead of doing two base stations make one and do beam forming.”

[The product] works, its live, it’s all there but the commercial success is not yet there. I think we’re too early but we decided we’d continue, go to second generation, make it cheaper and easier to operate.

This year we are bringing all our core elements as virtual [instances] on standard IT hardware, so the MME, gateways, HLR, you name it, all virtual and all on IT COTS hardware. We’re the first in the industry, nobody else is doing so, nobody else is putting the entire core on IT.

(To be fair, there are NFV and vEPC, vRNC, vIMS demos and announcements all over the show)

[Other telco vendors] are having some elements, but are still not using totally commercial COTS hardware, they are still on some kind of tailored ATCA hardware. So if you go totally to COTS, it means providing the resiliency for the non-breakable core, so you can build this functionality into the software. So we built this into the software when we virtualised it: not to break, not crash, even though the hardware might crash the software reconfigures automatically. This is what we are standing for – all the network on all COTS hardware.

Video on TCO reduction through NFV

On the PC every user has got AV and a firewall. Guess how many on a phone? It’s 3% – nothing.

For new LTE deployments we are now seeing network protection picking up, IP Sec key distribution we see now more and more. In total we have 500 security projects now running. 500 is a big number. EPlus is a big one in Germany. So the network security is getting there, that’s good to see.

But what still is not there is end user security. Now on the app store for Android there are 1.3 million apps that are infected with malware, and they sneak through the network to the phone. So the network is safer but the content gets more malware, and the end consumer not picking up on thinking, “I need to protect my device”.

On the PC every user has got AV and a firewall. Guess how many on a phone? It’s 3% – nothing. People think their operator is responsible, will take care of that, so said that we need to protect end users from inside the network.

What we’ve brought to the market is a solution together with F-Secure where we install a virus scanning engine in the network, not in real time data traffic flow but in control traffic so it is not blocking. On top of that we have developed algorithms that understand patterns of usage so if something sneaks through to the phone and the phone then behaves strangely, or suddenly a lot of phones generate a certain traffic pattern, we block that traffic.

So first we scan and clean, and then we kill traffic on the uplink and inform the user that the phone is doing something strange. We can block traffic to prevent bill-shock and offer the user some cleaning software.

So this is now live in a network in Europe, which is not public, and we are bringing it commercially to the full market by the middle of this year for operators to offer in a branded way. F-Secure provides the signature database and we do the behavioural part.

I’m convinced this is something that will pick up now in the market.

End user security video

Applied Big Data:
We invented the minority report for the mobile network because we’ve developed an algorithm, it’s applied big data, it’s really cool, predicting where an outage or service degradation in the network is happening in the next 48 hours with an accuracy of up to 95%.

The tool analyses weather conditions, traffic patterns, classic network KPIs, time of day, how many voice calls are in each cell, crunch it all together and we are able to predict that “in 48 hours you will have an outage there”. Then it’s about what an operator can do about it, reconfigure, remotely re-home some base stations, provision some capacity, define a cell differently. And then if that happens, and the outage is stopped before it happened, has it ever happened?

It’s not going to heal a really bad network but if you have a good enough network it can really predict service degradation and outages.

Predictive operations video



Leave a Reply