Rewriting economic equations of Internet ?Wednesday, July 25th, 2012 by Antonio Manzalini
We are moving towards an highly connected age, based on the complex interplay between social and ICT networks. Network Science will have a role in contributing to the challenge of rewriting economic equations of Internet. Actually, considering technology acceleration, it’s easy to imagine future telecommunication infrastructures be based on cloud ecosystems of virtual resources, with diversified s/w features and characteristics, capable of meeting dynamically requirements of next generation services. I think that borders between networks and data centers will blur progressively, or better a high capacity and “elastic” (i.e. self-reconfigurable) networks will be able to interconnect and serve ensembles of hybrid Virtual Data Centers.
The question is what are the key milestones to turn this vision into reality ? There are several technology trajectories under the spot. Software Defined Networking (SDN) is one of them, proposing the decoupling between s/w control plane (e.g. in charge of routing, traffic engineering, mobility) and data plane (e.g. in charge of forwarding the data packets) of future network nodes (e.g. based on general purpose hardware). SDN is not OpenFlow (which is today a protocol acting on decoupled forwarding tables) but it is often considered its future evolution. SDN should not be confused with Network Virtualization (which is bringing into network nodes the virtualization features normally adopted for IT resources), even if the two trajectories could intersect with interesting perspectives. Moreover it should be also noted that SDN s/w control is expected to be (at least today) logically centralized. Then we have autonomic, cognitive networking, etc … which are, to make it simple (filtering the hype), introducing algorithms, methods, rules, automatic control loops, and other Industrial Mathematics tools, into networks to make them capable of self-adaptation and self-management. From this short analysis, it appears that the common denominator is: decoupling software, more tools from Industrial Mathematics (to master complexity) and clouds of low-costs open nodes (by the way, this is reducing Opex costs, think about Google Data Centers). Are we ready for this change of mind ?
Clearly, this will create a ripple in the current markets, as it would mean, in principle, commoditizing (or at least opening) Vendors’ network nodes, and reducing dramatically network costs (both Opex and Capex); on the other hand this move would create the conditions for new ecosystems (with economy of scale) where Operators can gain ground in the QoS domain (i.e. programmability of services), even up to the edges, reinventing their biz in the Internet of Services.
In line with this reasoning, I’ve been captured by this piece of news, today. Technology-economics interactions are bidirectional: economic forces shape the evolution of technology, while disruptive technologies can rewrite economic equations.