Why Wireless Doesn’t Scale with Moore’s Law
Moore’s Law, which simply states that the capability/cost curve of chips should double every 18 months, hasn’t applied to mobile operators. Below are some reasons why the cost structure and performance/capabilities of wireless proves way more complex than silicon:
- Scarcity of wireless spectrum: Wireless spectrum allocation costs prescribe to a different cost/capability curve, one based off scarcity of supply. There is only so much wireless spectrum to go around (from a practical standpoint).
- Cell site acquisition costs: the complexities around allocating cell towers / sites do not scale well. The costs to acquire and deploy wireless cell sites is bound by real estate dynamics, government issues, and the like. Also new technologies are trending toward having microcells, instead of fewer macrocells, so deployment issues won’t get any better with new networks.
- Decoupling of data traffic and revenue: See ‘data era’ vs the voice era chart. Acute issue due to ‘unlimited buffet’ data plans. Wireless is quickly following the precedent set by the wired world (DSL, cable) of flat rate billing.
- No real substitutes for cellular wireless: Wi-Fi doesn’t offset the data deluge when people are truly mobile. The emerging WiMax will offer some relief, but won’t really make a dent for smartphones.
- It’s not just bandwidth: signaling issues are irrespective of bandwidth issues – big deal for apps like Foursquare and Twitter. Only a complete shift in network architecture will solve this (not more spectrum / bandwidth).
- Only as fast as the slowest bottleneck: like an Olympic relay, all of the racers can be fast but if one person slows down, everyone’s time suffers. The wireless networks still have bottlenecks in the access side – AT&T has furiously been adding to ‘backhaul’ to alleviate this slow spot.
it’s worth noting that fixed broadband (cable, DSL etc) hasn’t scaled with Moore’s Law either. We still have pretty slow home/office internet connections. Some of the same problems exist there, though the challenges are less technical and more to do with economics and laziness by the incumbents. Here I covered why Google’s announcement seems to indicate that some of the fixed broadband economic issues are going away.