Eulogy 💔 To a Hatchet 🪓— Reflections on “Accelerating Innovation in the Telecommunications Arena”

The Physics of Innovation with Aliens against the odds of #NFV

6 min readMar 30, 2022

A surreal scene from Men in Black (credit: Sony Pictures)

Back in 2020 a first draft of a paper called “Accelerating Innovation in the Telecommunications Arena” was published. It comes after nearly a decade of Network Functions Virtualization (NFV) in the wonderful world of Telecom, and it strikes me as an olive branch 🕊️ of sorts.

Unfortunately, the many years of contentious and strenuous coopetition have clearly taken their toll.

A cheek-turner on intent-based networking

The first page of the exec summary does not bode too well: It uses the word “dependent” nine(9) times when referring to other parts of the ecosystem, emphasizing a subordinate relationship that primarily exists in the minds of the authors. One might call this a “cheek-turner” — as a benevolent reader, I am frequently reminded to turn my other cheek as I am hit with a flurry of otherwise good intentions.

[gets on his soapbox] Every ecosystem represents a value network in which entities engage in transactions; every transaction is an exchange of something for something else, for example money (dollars) in exchange for a network connection or service. This implies that telecom companies depend on content and digital service providers too: they close the loop. A lightbulb 💡 won’t light up unless both wires are connected.[/end of rant]

One can never be sure, but I suspect the tone and attitude illustrated here constitute what Alan Quayle calls “a rather telco-centric view”

Innovation frequency model

The frequency illusion is a form of cognitive bias that causes humans to notice something more often after noticing it for the first time. Although the more common example is when you buy a new car, in my case this phenomenon may apply to two-phase liquid cooling solutions — please forgive me if that seems to be the case.

Reimagined Telecommunications Ecosystem Domains diagram

This modified diagram attempts to capture the notion that ‘innovation’ in the bottom layers is fundamentally different from ‘innovation’ at the top. Apart from the increase in cycle rate (iteration frequency), the process is qualitatively different at each layer, like a phase change in the case of liquid cooling. As P. W. Anderson once said: “More is different”

This 2019 article on “Why move fast and break things doesn’t work anymore” suggests that until recently, the innovation paradigm was focused on rapid iteration, a.k.a. “move fast and break things”. Netflix’ Chaos Monkey illustrates one of many tools and approaches that “Digital Services Providers” use to keep up with the rapid pace of changes. However, like all things, this paradigm will eventually reach a limit, and a different approach that is “more slow and deliberate, integrated across multiple domains” may take over.

The analogy with heating/cooling cycles also extends to levels of excitement and agitation of the humans involved. Although I am no psychologist or neurologist, some actors involved in the faster moving parts of the eco-system seem to experience higher levels of stress than others, which may have negative side effects. Without calling out or implying anyone in particular, I believe this “flip side” of rapid innovation contributes to the complex dynamics.

The definition of ‘innovation’

The paper uses the term ‘innovation’ 30 times, without including a definition of what it is (or assumed to be). According to Wikipedia,

Innovation is the practical implementation of ideas that result in the introduction of new goods or services or improvement in offering goods or services

However, it is hard to quantify this concept: What constitutes a single idea for example? If a Service Provider introduces a novel pricing scheme, is that “one” innovation? What if the scheme consists of multiple discounted bundles? What if they use AI to segment the market? etc. etc.

In spite of the missing definition and quantification, there is an implied notion that — somehow — innovation must be “accelerated” for Telcos (see title). So whatever it is, we need more of it (RFC1925 point 5)

Innovation requires a problem, Telco networks provide several

In order to drive innovation, there must first be a problem to be solved. Few people do innovation for the sake of innovation; even if government subsidies are involved, ‘adhering to the conditions with minimal effort’ could become the problem to solve.

Around 2009 Skype first introduced and later open sourced its SILK codec: An advanced adaptive audio encoding technology that was capable of dealing with highly varying network conditions such as high packet loss, jitter and varying bandwidth availability.

The driver for this innovation? Complaints from early adopters due to the issues with unreliable telecommunication networks. The network triggered this innovation, but Skype did not succeed because of the reliable network — it succeeded in spite of the network. Skype may have indirectly been helped by the network, because it stood out among its competitors by the way it handled network quality issues — I’m not sure. In any case, Teams took over and Skype is being phased out, having completed its mission.

In a way, barriers are enablers of innovation — they create pressure and friction, and play a role in natural selection and elimination of viable businesses.

Metrics for sustainable growth — is “innovation” the goal?

When judged by the perceived rate of change as communicated in press releases and social media activity, “Digital Service Providers” and “smaller new players” could be said to be “more innovative”. Although not stated explicitly, the paper seems to “glorify” a (vaguely defined) notion called “innovation”, implied by questions like “how to remove barriers” and “how to encourage more”.

Is it truly “innovation” that is lacking here? I don’t have the answer, but my question to the authors would be: What problem(s) are we really trying to solve, for whom and for what purpose? What are we trying to achieve? Can we identify and agree on the metrics that one might target, using an agile process?

Final thoughts

Opinions differ on the merits and degree of success for “NFV”, but reaching consensus on that point would not solve anything. One cannot change the past, but we can learn from it, try to understand what happened and hopefully avoid some of the negatives and achieve more of the positives.

As we bury our hatchets, I leave you with some closing questions to consider:

✅ Do devices like iPhone and Android smartphones not provide “ubiquitous connectivity and access to the global internet”?

✅ Considering this new mobile service edge, to what degree has “softwarization already changed the way that telecommunications infrastructures are constructed”?

✅ Who are the physical infrastructure providers? Have they really delegated standardization to their suppliers?