7 Comments

Summary:

Written by John Roese, Chief Technology Officer, Nortel Networks Is 4G about improving the performance of today’s mobile networks, or is it about revolutionizing the model to create a truly ultra-broadband mobile experience? The current dialogue is focused on improvement, notably how to improve the capacity […]

Written by John Roese, Chief Technology Officer, Nortel Networks

Is 4G about improving the performance of today’s mobile networks, or is it about revolutionizing the model to create a truly ultra-broadband mobile experience?

The current dialogue is focused on improvement, notably how to improve the capacity of the cellular infrastructure via such 3G technology iterations as HSDPA/HSUPA and CDMA 1xEV-DO Rev A, or such revolutionary 4G technologies as mobile WiMAX 802.16e and LTE.

But simply improving the capacity of 2G networks will not be enough to create the kind of network experience the hyperconnected world expects (see my blog for more details on hyperconnectivity). In order to mobilize the Internet, we need to look not just at capacity but at the
experience that the current mobile network has, thus far, not been designed to deliver. Once we do that, it quickly becomes clear that achieving true Internet mobilization will require more than just minor enhancements to the existing mobile network — hence the need for 4G.

The three key features of the next network include:

1) Support for unsubsidized end points: The network must allow for the connection of any device that would benefit by being attached. Phones, laptops, MP3 players, security cameras, automobiles, sensor systems — all of these and more must be considered potential consumers of mobility services.

2) Open applications ecosystems: The 4G network has to be designed so that it can transport any application that would benefit from a mobile broadband experience, allowing for an infinite set of applications to become participants in the mobile ecosystem.

3) Dramatic economic improvement: The system must facilitate a cost model that focuses on the volume of users and devices — not on higher ARPU for a small set of defined devices. A lower per-bit cost per is important in 4G, but so too is a lower cost for provisioning, operations, backhaul and service abstraction.

The existing 2G and 3G networks are purposed to deliver voice, narrowband data and specific video services in a structured, well-defined model. The 4G network, meanwhile, needs to be designed to mobilize broadband Internet services much in the same way as home and enterprise networks. So while these systems overlap to some degree, the inherent purpose of each is different enough to warrant new technology and new assumptions.

Over the next five years, there will be both improvements made to the existing cellular experience as well as the introduction of 4G systems that expand mobility to a broader set of users, devices, applications and customers. And as the 3G and 4G worlds evolve, they will intersect and converge. The key for us as an industry is to recognize that the short- and medium-term coexistence of long-term competing technology is normal — and natural — and that rather than resisting this evolution, we should embrace it.

  1. Krishna Chodavarapu Monday, September 17, 2007

    It is indeed exciting that several technology companies have realized the need to wrestle the control of the country’s infrastructure away from the telecommunications industry. Perhaps it is reality, or only their online caricature, but it is my long standing belief that these dinosaurs are not consumer friendly and believe the airwaves and the “pipes” to be their birthright. It is time for an agile group of mammals to put them in their place.

    I hope that you are sincere in calling for an investment in 4G infrastructure. I also hope you will partner with the likes of Google, Sprint, and others (AAPL?) to take on the more established and deep-pocketed AT&T and Verizon.

    No doubt this calls for a large investment; but I strongly believe consumers will reward you for it.

    Share
  2. [...] Networks CTO, John Roese, penned a post on GigaOm about the possibilities of 4G. He point out flatly that “simply improving the capacity of 2G [...]

    Share
  3. Good points. The fact that capacity improvements alone will do little or no good has been long known; even currently the limitations of mobile service usage have nothing to do with network capacity or even performance. To be blunt, the industry tunnel vision of focusing only on performance is, if not entirely, at least largely misplaced.

    Of the three key features you mention, many of the current mobile networks already support #1. Many are also approaching the open applications environment (#2), whether the operators really want that or not. The economic improvement, however, will take time – and may not materialize in quite the predicted forms. The network/service provider split and the ensuing rise of pure mobile connectivity providers will, however, help.

    Another thing is that we should not call whatever the future will bring “4G”; calling a gradual development and a fundamental change of the industry including economic structures with the same terminology as straightforward performance improvements earlier is misleading at best.

    Finally, I don’t want to sound overly pessimistic, but five years is a pretty tall order for anything even remotely resembling 4G to appear for the general public to use. Remember Amara’s law.

    Share
  4. [...] is beckoning a bit beyond the horizon. The question, which is nicely posed at GigaOm by John Roese, Nortel’s CTO, is what precisely 4G will be. The post doesn’t dive into [...]

    Share
  5. Ha! 4g?? We don’t even have 3g yet in India… need 3G…

    Know how I feel? Check out India Web 2.0 News for Indian tech related news!

    Share
  6. [...] year or so ago, I began talking about how moving toward the 4G world of WiMAX and LTE (and UMB at that time) would enable a broad range of new devices to be connected economically to [...]

    Share

Comments have been disabled for this post