When Van Jacobson has something to say, people tend to listen.
His algorithms for the Transmission Control Protocol (TCP) helped solve the problem of internet congestion and enabled it to survive a major traffic surge in 1988-89 without collapsing.
Jacobson (pictured) is currently a research fellow at PARC, the company that invented ethernet, and is warning telcos that new problems lie on the horizon.
In short, he says the architecture of the internet needs to change to cope with its transformation from an end-to-end communications network into a distribution network for multimedia sharing.
Thanks to the strain on their networks, telcos are well aware of the content revolution underway; however, Jacobson said the "fixes" they have used to handle the shift will not last forever.
“Every day millions of dollars of capex and opex are invested in tricking [the internet] into operating far outside its design parameters,” said Jacobson in a statement launching the Emerging Network Consortium last week.
“Internet-based content consumption is reaching a scale where tricks no longer work.”
The ENC has brought together a range of partners to work on PARC’s answer to this problem: content-centric networking.
CCN is a network architecture that directly routes and delivers named pieces of content at the packet level of the network, enabling automatic and application-neutral caching in memory wherever it’s located in the network.
Designed to run alongside or independent of TCP/IP and not to disrupt existing networks, PARC claims CCN will enable efficient and effective delivery of content wherever and whenever it is needed.
Arjun Nandal, head of solution design at MACH, one of the founding members of ENC, told European Communications that it would enable operators to fight back against the OTT players.
“CCN is a natural way for operators to evolve their business models through value-added services and move away from dumb pipes,” he said.
Both BT and France-Telecom Orange are also members of the ENC, but in briefings with European Communications took a more cautious approach.
Keith Blythe, head of future consumer applications & services at BT, said when conversations with PARC began, the UK-based operator thought their proposal was too radical.
“Since then our ideas have converged – it’s now evolution not revolution,” he said. “However, lots of questions remain, particularly over costs and whether it will work in practice, so it’s too early to call at present.”
Prosper Chemouil, R&D programme director at FT-Orange, said while CCN could certainly strengthen the hand of operators, it was just one of many approaches to the problem.
Both operators were in agreement that their “fixes” were able to deliver what they needed at present.
Nevertheless, they were keen to explore whether CCN could deliver on its promises in the future – Chemouil said it was unlikely to be fully deployed before 2020.
Its success is likely to be dependent on a group that is conspicuous by its absence from the ENC – the content providers.
Chemouil said the biggest challenge remained agreeing new business models with parties such as Google.
Clearly, if better deals than the ones that currently exist are unable to be struck then it doesn’t matter how good the architecture/technology is, operators won’t be much better off.
Understandably, the vendor community is much more upbeat, sensing an opportunity for a new market.
The ENC lists Huawei among its members and John Roese, a SVP at the China-based firm, said CCN was “disruptive thinking and entirely necessary work”.
One group who might be less happy with proceedings are providers of content delivery networks, which are one of the “fixes” that operators are using at present.
Perhaps it is with this is mind that Jatinder Singh, director of mobile innovation strategy at PARC, said the biggest challenge was in helping the different parties to come together around CCN.
Officially less than a week old, it is clearly early days.
However, as BT’s Blythe admitted, the ENC warrants attention because of the significant parties it has brought together – in addition to the companies mentioned above, Alcatel-Lucent, Panasonic and Samsung are also involved.
If CCN can be proved – and with Jacobson’s pedigree you’d expect it could – then attention will turn, once again, to content providers.
Operators may well have some extra ammunition to bring to the negotiating table, but given the long-term timescale the question has to be asked: will it be too little too late?
Have you taken our new Big Data Survey? Click here to share your thoughts.