Transformative network analytics must see beyond basic usage patterns.
Data is the fuel behind the power of digital technologies. Without it, no statistical models can be trained, no innovative applications can be built and, ultimately, no new knowledge can be leveraged for concrete value creation.
While native internet players have mastered the ability to turn digital data into business-empowering insights, the Communications Service Providers (CSPs) responsible for enabling all that data exchange in the first place have done little in the way of producing genuinely new digital-centric intelligence.
CSPs can, nevertheless, still strive to speak novel languages. It remains within their reach to equip themselves with deeper understanding over the experiences their services enable and, through that, become more relevant in the digital knowledge generation spaces.
“The fundamental characteristic of a knowledge plane is its ability to ‘see beyond’, grasping more events and sensing more contexts than old-school user plane.”
A great deal of that opportunity stems from novel network analytics technologies. Initially developed as tools to only monitor networks’ control plane and gauge their technical performances, those solutions have gradually incorporated the ability to detect and classify user plane events that shed light on collective online behaviours.
Nowadays, network analytics solutions have come to offer the only plausible alternative to understanding our worlds to what companies like Google, Facebook or Baidu already obtain by themselves. Those solutions open up to CSPs, then, a new level of intelligence generation: it gives them access to a knowledge plane.
The fundamental characteristic of a knowledge plane is its ability to ‘see beyond’: to grasp more events and to sense more contexts than old-school user plane can. That evolved visibility hinges on a superior ability to interpret data traffic, one that surfaces multi-dimensional data attributes carefully weaved together.
In Niometrics, we have strived to expand our Deep Network Analytics (DNA) platform with tailor-made components for knowledge plane development. Much of that effort has stemmed from the coupling of SLANG (our proprietary signature language) with NCORE (our deep packet inspection module). Together, they accurately identify encrypted applications and decode industry network protocols.
Of particular relevance for a proper knowledge plane rendering is the ability to detect application components: in-principle encrypted traffic which can be labelled with more complex metadata heuristics. By doing that, we can – for example – discriminate between a series of intra-Facebook actions: a message being sent, a picture being uploaded, a voice call (through Messenger) being made, and so on. Since nowadays a considerable share of online behaviour unfolds inside the properties of internet players, components can bring about a powerful nuancing to knowledge plane depictions.
With the right confidentiality measures in place (permission for, anonymisation and aggregation of individual data points being the non-negotiable ones), knowledge plane ready technology can place in the hands of CSPs a revolutionary capability to extract true ‘meaning’ out of all the data they transport.
The concrete applications for that type of intelligence are varied. Some are immediate (i.e. to transform how CSPs themselves serve and relate with their clients) while others may need bolder business model experimentation to crystallise. None of them, however, stands a chance of creating distinctive value if it doesn’t build upon unique insights over digital behaviours, exactly as knowledge planes embolden us to do.