Showing posts with label mit. Show all posts
Showing posts with label mit. Show all posts

Saturday, March 17, 2007

Public speaking: MIT CNC -- The Future of Portable Communication

I was invited to be a panelist at the MIT Club of Northern California's Entrepreneurship series "What’s In Your Pocket? The Future of Portable Communication" event. Other panelists were from OQO, Digital Chocolate, NVIDIA, and HP Labs. Julie Ask from JupiterKagen moderated.

My fellow panelist Susie Wee from HP Labs wrote about the event in her blog. She quoted me as expressing some reservations about the viability of WiMAX in the face of a nearly ubiquitous Wi-Fi install base. This is essentially correct, however WiMAX being a somewhat diaphonous term it's important to make distinctions on what use-model I was referring to.

I personally believe that Mobile WiMAX (IEEE 802.16e) will never happen due primarily to the nearly universal attach rate of Wi-Fi to mobile consumer devices, which severely drives down the potential for widespread adoption of Mobile WiMAX. It's estimated that by late 2009 it will be essentially impossible to purchase a laptop without 802.11, and that most will be 802.11n. This trend likely holds true for handheld devices (PDAs, SmartPhones, etc) if for no other reason than chip manufacturers will simply stop building non-11n parts at some point.

The essential question is; If the installed base technology is viable, and a network exists, then why change? Novarum has done extensive real-world use model system testing, and subsequent analyses in their "Wireless Broadband Rankings" and "Metro Wi-Fi Rankings" show that actual throughput of many municipal Wi-Fi deployments is approximately the promised performance of as-yet-undeployed WiMAX systems. 3G & LTE deployments from cellular carriers are increasing footprints daily and coming up the throughput curve. So why change, and why will anyone invest capital in spectrum, equipment, siting, and deployment of a Mobile WiMAX system that in the end will only be as good as what's already available?

The logistical and financial challenges of deploying wide-area wireless networks are rather severe and (most fortunately for those people trying to get investors to drink the WiMAX Kool-Aid) not well understood by more than the relative handful of people who've actually done such work. Those who don't understand history (anyone remember Metricom's Ricochet network?) are doomed to repeat it.

I believe that Mobile WiMAX might find a niche in non-stationary backhaul i.e. broadband on commuter trains, etc but I don't believe it will ever be the "last yard" connection for mobile devices. Fixed WiMAX has a better chance, especially in rural areas of the US which aren't served easily by broadband. Fixed WiMAX also stands a strong chance of succeeding in developing regions outside the United States.