Showing posts with label mobiTV. Show all posts
Showing posts with label mobiTV. Show all posts

Thursday, November 8, 2007

What Android Must Do


"I think for the Google platform to really be a game-changer it's going to have to offer more than just an open-source operating system for a mobile phone," says Kay Johansson, MobiTV CEO. "It will have to create mobile Internet devices that happen to make phone calls."

If Google ever does decide to bid for 700 MHz spectrum, with or without other partners, the network will be shaped by that same requirement. In fact, Google might have an opportunity to consider a different approach to what a mobile network looks like. Such a network theoretically might provide broadband coverage anyplace people are walking, spottier coverage in other places. That was the original thinking behind the "Personal Communications Service" (PCS) spectrum that instead wound up being deployed as 2-GHz cellular instead.

That might wind up being what happens again. Today, it is the troubled muni Wi-Fi approach that most resembles the old PCS idea. The Fon, T-Mobile Hotspot or Boingo hotspot approach offers less coverage than the original PCS concept called for.

In fact, the only reason one would logically want to build a new network on anything other than macrocells, incorporating Wi-Fi in some way, is that it might be possible to get a network up and running faster than if a traditional macrocell approach is used. One already would begin with fairly ubiquitous coverage in homes and offices and a reasonable hotspot overlay. Macrocell coverage probably still would be needed for outdoor coverage.

Still, an ad-driven business model ultimately could be quite do-able if one generally assumed in-home and at-office coverage, with fairly dense coverage in downtown cores and shopping malls, augmented by lighter coverage other places. There are some drawbacks, especially when the device is used in phone mode, since coverage would tend to be the reverse of what they now encounter. That is to say, indoor coverage would be better than outdoor coverage.

The point is that if the exercise is to build an optimal mobile Web network, one might have different choices. One of the bigger advantages would come from crafting a network that could be built at far-less capital intensity than a typical 3G mobile network. That would fit with the general theme of creating less-expensive handsets and service as well, so ads would produce a healthy revenue model. It's just a thought.

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...