Have a question for us? Use our contact sales form:
It is entirely conceivable that the future of Cloud Computing is so distributed and ethereal as to be completely indistinguishable from the network itself.
We can already see the shift in reality as wireless carriers struggle to keep up with the resounding demand in network bandwidth. As the old saying goes; 'Necessity is the Mother of Invention.'
In a recent statement, John Donavan, AT&T CTO, points out the futility of Moore's Law, lamenting the fact that while transistors may continue to double every 18 months as predicted so long ago by Mr. Gordon Moore, mobile network demand is growing at a pace almost twice that.
So how does one grow network capacity at a rate that is faster than the processing performance that is required to enable it? You go parallel.
The content distribution network people have already reached this conclusion, distributing their most popular content to web caching servers strategically located in the distribution network. So, how far away is the possibility that the mobile carriers themselves will join the ranks of the cloud computing culture? In fact, what is to stop them from infusing that cloud computing power into the Radio Access network (RAN) itself?
You can see it already. The world's ultimate cloud computing concept has been with us for years now in the form of Peer-to-Peer networking. While enabling P2P may not be in the mobile carriers best interests from a monetization standpoint, they could adopt the next best thing by distributing content into the GGSN, RNC, Base stations (a fairly harsh environment), or even... people's femtocells. Sure. Why not? The carrier could eat the cost of equipping everybody's femtocells with terabyte hard drives, large portions of which are reserved for caching carrier content. The cost would be insignificant compared to the challenge of backhauling a wireless network whose demand for capacity is growing at close to 130%. Even placing it at the regional data center would be a huge improvement over backhauling this huge volume of traffic. We can envision the X2 handover protocol in LTE networks facilitating just this sort of distribution.
Imagine if every femtocell came equipped with a 1 TB hard drive, of which half was allocated to the consumer and half to the network? In a 100 femtocell neighborhood, that is enough storage capacity for 50TB or 1000 double-layered Blue Ray discs. That is a fair amount of cached data.
None of this will help with the RAN capacity mind you, but it just might reduce the infrastructure costs of backhauling Exabytes of data.