Oh you know, I can’t prove it because this is my first bit of spouting off in this space. But if I had had another life before this one, say a website where I was occasionally allowed time off coal-mining duties to write something, you might have seen me once before, or even twice, mention that automation of network processes is going to be a big thing.
I even tried to write an article about it in our launch issue of TMN Quarterly , and although I’m not sure I quite got to the nub of what the vendors are up to, I’m gratified that my hunch that automation would be a “thing” right about now is being born out by announcements made after I went to press on that article (no, really, great timing everyone).
I digress. What do I mean by “automation of network processes”? I mean the automation of designing and optimising a network, tweaking and operating it so that it works well and delivers the best performance at the right time in the right place. Things like that have relied on clever people doing clever things up to now, and now they will increasingly be done by clever computers doing clever things.
Partly this is being driven by the Het Net, with its demands for inter-layer coordination, and for inter-mode traffic management (put simply, managing services across 2G, 3G, and LTE networks). We might call that the outside-in optimisation – where things happening in the RAN are fed back to the core, whicih decides what to to about them. But I also think there is a movement that is moving from back in the network, right in the back office, out to the edge – let’s call that inside-out optimisation. This is where customer and network data is being churned through to produce actions such as tilting an antenna, yes, but also policies and classes of service.
So, you’ve got SON, yes. But I think SON is morphing, as a term, from something specific (self-configuration, self-healing, self-optimising, standardised in 3GPP) to mean, “automatic things that happen in a network that take advantage of data that we have got from somewhere”.
I think that’s happening because the main vendors have realised that they need it to mean that. With the outside-in model, there is potential disruption to their hegemony from stand-alone planning and optimisation companies, doing clever things right from the device level to inform the network what is going on, and what should happen as a result. With the inside-out model, disruption can come from the likes of a Tektronix or big OSS vendor, or someone who can crack the “big data” thing. (I like calling things a “thing”. It is suitably vague and keeps us moving on).
That’s why we saw Ericsson buy Telcordia and why NSN made its CEM platform, built lest we forget on the long-forgotten acquisition of SDM specialist Apertio in 2008, a central plank of its bid to be the world’s most focussed Mobile Broadband Solutions company.
Am I making this stuff up? I don’t think so.
Am I making this stuff up? I don’t think so. Here’s NSN making its pre-MWC push. Capitals are mine. Look through the announcements and note…Network 360 – a map view of the network to SPEED UP troubleshooting… AUTOMATED LTE Site Creation… AUTOMATED subscriber balancing of MSCs… “our goal is to help operators improve information flows, AUTOMATE their systems and processes, and prevent the isolation we often see between many functions in the operator organisation”
Here’s Ericsson’s equivalent: Operators
Look, some of this focus on automated processes, from RAN to core to policy etc etc is in response to operator demand, and some is in response to the vendors’ own positions. Ericsson has previously argued for tight inter-layer co-ordination between small cells and the macro layer – something that would best be done in a single vendor environment within a given sector: it’s vertically integrated from its OSS through to the radio layer – end to end, if you like. Conversely, you’ll notice NSN talk more about managing multi-vendor environments: it is less integrated, but still wants a role in services and in smoothing acceptance for its LiquidNet roadmap.
If I were Aircom, or Actix, or Arieso or Amdocs or any other company that begins with A and plays in this automated orchestration and optimisation area, I’d be pretty happy about this. It’s validation of what they’ve been saying for a while now – that network smarts, whether distributed or centralised, or a bit of both – are key to network economics, especially in LTE.
At MWC, we should look out for these automated, automatic messages, and remember why we’re hearing them, and who we’re hearing them from.