Sunday, October 22, 2006

Phone 2020 - disrupting business models

The combination of flexible use spectrum licenses, software-defined radio, and mesh architecture could be a disruptive change in the next decade:

We’re just at the threshold of the new spectrum licensing regime. We’ve been seeing spectrum prices trend down, and that’s likely to continue particularly with more spectrum in flexible use. I think few people really understand what impact this regime will have effective spectrum availability

SDR combined with lots of places where the radio can operate will blur the boundary between licensed and unlicensed, and between PAN, LAN and WAN services. As far as users are concerned, it’ll be “just wireless”. SDR as such isn’t all that special; multiple cheap radios are already pointing the way. However, SDR will make system integration a little easier – provided that the power/compute challenges are met.

We’re asymptotically approaching the Shannon limit for single channel communications. However, multi-user information theory is still an active research field. It may turn out that Shepherd, Reed et al. were right that one can create architectures where capacity scales linearly or better with the number of nodes. (The best results I’ve seen so far scales with the square root of the number of nodes, so per-node capacity still decreases.) If this dream comes true, bandwidth per bandwidth (bits per hertz) could be much higher than current prudent planning assumes.

If you add these three together, the business of cellcos changes dramatically. Spectrum scarcity as an input constraint will be considerably weakened, as will the cellcos’ balance sheets – typically half the value of such companies is their spectrum licenses. If cellcos are to survive in this scenario, they’ll have to change their business; for comparison, once anyone could access airline ticket databases, travel agents could no longer live on commissions alone.

The change is analogous to the experience with watches. A hundred years ago, one paid for a watch in order to buy a timekeeper. Nowadays, we pay either for style (luxury watches), or for time management (calendar software, personal organizers).

Today we still buy communications when we pay for a cellphone; in fact, we’re buying the right to use a sliver of spectrum. In the future we’ll either buy style (designer phones) or comms management.

Communications management is required at various network layers. To simplify simply, we need comms management at the transport layer and the experience layer.

Transport management is a future business for cellcos: seamlessly combining personal, local and wide area networks into a cheap, high quality communications service. T-Mobile’s combined offer of telephony over Wi-Fi and cellular networks is an early incarnation of such a service.

Experience management is the business of content portals. MySpace is today’s paradigmatic example. Teens use text messaging and MySpace to stay in touch – not email. Places like MySpace provide the social structure which anchors the concentric conversations (1-1, communal, ambient) that will replace phone calls.

One can describe a comms management provider strategy as a “tuple play,” by analogy to the “triple play” and “quadruple play” in consumer broadband access. The goals is to integrate multiple communications services (the tuple) into a single product.

No comments: