The End of Gs
Finally back on California time, after a busy week in Barcelona at the Mobile World Congress. The GSM Association, which owns the event, changed the name from the 3GSM World Congress, dropping the reference to 3G. In part, this represents the desire to attract participants from the information technology, financial services and entertainment industries (including Robert Redford). But dropping references to radio technology (2G GSM or 3G W-CDMA) also suggests the “end of Gs” for the mobile industry. Not that new radio technology was ignored at the Congress. Many heralded the arrival of 3GPP Long-Term Evolution radios (although the impression of imminence is undermined by the LT in LTE), others exhibited WiMax (including Cisco), and femtocells generated much interest (including partner ip.access). But the notion of 4G replacing 3G, which replaced 2G, which replaced 1G, makes decreasing sense as a framework to describe the future.Past evolution was driven by changes in radio technology, from analog to digital and then to CDMA. As the radio technology changed, the standards bodies (3GPP and 3GPP2) also changed the core. But future radio evolution is unlikely to come in tidy generations. Instead, mobile networks will use an increasing blend of radio access. The ongoing evolution is now driven by a change in dominant payload, shifting from voice to data, with traffic shifting even faster than revenue. Data demands ever higher bandwidth. Fortunately, Moore’s Law will unleash ever faster radios (and in more combinations), but business cases will spread their adoption over time, while history and regulation spread their adoption across geographies. Also, physics and economics dictates different radios for different applications, for slower speeds over longer distances, like cellular, and higher speeds over shorter distances, like WiFi. Thus, mobile operators should expect not a monochromatic shift from 3G to 4G, with a simultaneous re-engineering of the core network. Instead, they should plan for a rich palette of radio access, with a core comprising a network of these access networks, that is, a core mobile internetwork. This change explains why so many mobile operators increasingly consider IP the foundation, around which to array the growing collection of radio technology. New radios should have no more effect on architecture and applications, than upgrading links from shared 10 Mbps to 1 Gbps Ethernet. Of course, this focus on the mobile internet and IP is exciting for Cisco, but it also raises the challenge of accelerating our mobile gateways to meet the traffic growth. At the Congress, we demonstrated the SAMI platform for mobile gateways (GGSN, PDSN, content services), which Current Analysis recently tested up to 54 Gbps, making it the fastest in the industry by a factor of more than five times. With the growth in traffic from the iPhone and other devices, it comes into service just in time. As always, Barcelona was full of beauty and fun. In a park near the Congress, I came across a huge outdoor sculpture by Joan Mira³, called “Dona i Ocell” (“Woman and Bird”). While photographing it with my camera phone, I noticed the pigeons behind me squawking very strangely. Only they weren’t pigeons, they were parrots living in the palm trees. More international arrivals, along with the Mobile World Congress participants, all trying the patience of the locals with our chatter.