Can wireless internet ever match fixed service performance? The wireless advocates say that their technology is fast closing the gap while the naysayers argue the opposite—wireless nets simply can’t handle the demand when you place mass market loads on their networks.
Both sides can point to hearsay and anecdotal evidence to back their points of view. The wireless defenders point to the marketplace, for example, the recent ABS stats which show massive growth in wireless net connections to the point where 1 in 9 Internet subscriptions are now wireless. But the fixed advocates say that even at that level, wireless networks are straining, pointing to recent reports that Optus’ 3G network is struggling to handle user loads as evidenced by its withdrawal of popular plans from the market.
It’s more than a technical geek debate—it impacts some of the great issues of our time. For example, much of the regulation of Telstra—which goes to the heart of its complaints about network investment obstacles—stems from the assumption that its fixed network constitutes a monopoly unassailable by wireless. And likewise, the Federal government’s request for FTTN and FTTH proposals seems tacitly biased against wireless services, even though under the last government the latter were seen as the most economic solution to rural broadband market failure.
In an ideal world, one would simply submit the rival platforms to a Mythbusters-style face-off– how do they stack up in real world conditions? But even where this has happened the results seem somewhat confused and inconclusive.
A case in point: two of Australia’s major computer magazines this month offer cover stories on the results of their own wireless broadband tests. The surveys offer diametrically opposed results.
Australian PC Authority approached the task from a holistic point of view, not just looking at data download speeds but also the performance of customer equipment such as HSPA cards and modems as well as the pricing packages offered by the various providers. Its final verdict represents a blend of technical performance and pricing value.
Its verdict? Hutchison’s 3 is the best provider, followed by Optus, which was “recommended”, with Telstra, Vodafone, Virgin and Unwired bringing up the rear.
Australian Personal Computer approached the same task in a slightly different way. It focused more on technical performance, comparing how the 3G services compared on a variety of set tasks—downloading specific web pages, streaming specific videos and generalised speed & latency tests.
Its verdict? Telstra Next G the hands down winner, followed by 3. Optus, Vodafone and Virgin were consigned to the also-ran category. The APC survey was particularly scathing of Optus/Virgin.
THE DSL CONTROL: One of the most interesting aspects of the APC tests was that Next G was closer in performance to a control ADSL2+ service than to the other 3G services. For example, Telstra Next G took 10 seconds to load the Lonely Planet webpage compared to the ADSL2+ performance of 9 seconds. The same webpage took as long as 45 seconds to load over other 3G services. But as I say, this is all interesting but not conclusive. For example, the ADSL2+ control service employed by APC was sourced from Internode –a choice made by the magazine I suspect because of that firm’s reputation as the independent carrier of choice for high-end power Internet users.
But surveys, being what they are, can play tricks with market perceptions. Last week, Internode was ranked dead last in an audit of broadband performance across the major ISPs, compiled by Epitiro Technologies (albeit unavailable publicly but presented at our CommsDay Melbourne Congress last week). As you would expect, Internode was flabbergasted by this result, given its whole business proposition is based on superior technical performance. But the survey methodology appears to place Internode below the industry average on 6 out of 8 parameters, despite market evidence that Internode’s power users are happy with their lot.
Unfortunately all this leads one to a depressing conclusion—that even the most noble-minded and well-intentioned independent testing of the respective merits of fixed and wireless technologies will never prove conclusive. Which means that some leaps of faith might need to be taken.
I learned last week that one of the major proposals for the national NBN process will call for FTTN to 93.7% of the population and LTE– the next upgrade to HSPA—for the remaining 4.3%. The NBN Expert Panel will be compelled to make a call on this whole fixed vs wireless debate.
Grahame Lynch
