Over the past few months numerous press releases have touted so-called "speed tests" performed on early LTE networks. At first it was easy to dismiss these announcements of spectacular performance as little more than the usual new technology hype.
However, as the number of reports indicating exceptional peak speeds and extremely short latencies have started to pile up, we need to consider the implications of what truly could be game-changing performance.
One of the latest and more credible tests was published by Epitiro, a UK-based firm of broadband "quality of experience" analysts. In a fairly large test conducted over a period of one week and resulting in 20,000 data points, Epitiro found that TeliaSonera's LTE network in Turku, Finland delivered an average download speed of 36.1 Mbps and just as importantly delivered an average latency of 23 milliseconds, with little delay variation. In addition, it measured upload speeds of 1.7 Mbps over the LTE network.
All of these metrics improved on the performance of the parallel 3G network by factors of five to ten. We can discuss the relative merits of such testing at length - it was conducted on an unloaded network for instance, from fixed locations, etc. But the conclusion is actually fairly clear: under good conditions, LTE will deliver performance that blows the pants off "regular" HSPA (not necessarily the many flavors of HSPA+), and Wimax for good measure.
The data rate is only one issue. Given enough spectrum and similar MIMO configurations, all of these technologies should deliver similar speeds. However, LTE is delivering much better delay performance than both HSPA and Wimax, and much better uplink speeds at least compared to HSPA. These factors have a direct effect on user perception and the range of applications the network can handle.
Clearwire CTO John Saw, presumably someone who has conducted some pretty serious testing, praised LTE in a recent interview not only for its outstanding speed using large channels, but also for maintaining that data rate at vehicular speeds and during handovers.
So, enough about performance. The interesting question is how will this change the industry? Basically we've got a technology on our hands that in its preliminary state seems ready to blow the socks of 3G, and do a creditable job of providing residential broadband services.