Log in

View Full Version : 3G's poor handset performance



3g-g
7th May 2005, 05:42 PM
An intersting piece from el Reg (http://www.theregister.co.uk/2005/05/07/bad_3g_phones_need_proper_testing/) , it's seems there's mounting preasure on the HS manufacturers to make 3G more attractive to users in terms of signal and coverage compared with GSM. Do the people who write these things understand that it's a completely different technology... that's only just rolling out at that!


The 3G phones in the market are not up to snuff, according to experts: "We are seeing performance differences between two 3G terminals, with as much as 6dB or 7 dB falloff in signal sensitivity. Inside a building where signals are marginal, that's the difference between 'working' and 'not working' - and it's not being caught in testing."

The accusation comes from the leader in terminal (phone and handset and data cards, to us users) testing: Spirent. Director of Applications Engineering, Nigel Wright, said frankly: "The handset makers have concentrated on passing 'conformance' tests, at the expense of real-world performance."

Unlike the standard GSM cellphone in Europe a 3G handset is hard to test.

The "time slot" system of sharing antennae on GSM is replaced by a CDMA contention management system, where the number of users on a base station can vary widely.

On Wideband CDMA (3G in Europe is all WCDMA) the critical problem is power management, both in the handset, and in the base station.

"The handset has to manage power, because of the need to maximise battery life," said Wright. "But it's not a simple equation. The more users in a cell, the more power you have to pump into the signal, and the less power you pump in, the more errors users will see on the fringes."

Users of 3G data services will all be well aware of the problems of trying to get error-free communications, especially on data cards. But it's very difficult to set up a good real-world test for phones or data cards which is repeatable, and as a result, the requirements of UMTS "conformance" tests have become very watered down, Spirent thinks.

"We're well aware that a few big handset makers have put pressure on the committees to allow performance which isn't optimised for anything except passing the conformance test," said Wright. "It's a problem which the networks are very concerned about, because users are seeing calls dropped, errors, and loss of signal, which the networks are blamed for. But the problem is in the mobile terminal, in our judgement, in far too many cases."

Hands0n
8th May 2005, 07:12 PM
I find myself in general agreement with the sentiment of the article. I have been a professional (corporate) and private "early adopter" of technologies for lets just say a few years (lol!). The excuse traipsed out for the early days/weeks/months/years of poor performing systems and services is the same. Of course, I do understand that there are very high commercial pressures on those suppliers providing the services to get these to market very quickly. These pressures often result in the product being released in a somewhat raw state. The ensuing days/weeks/months/years are then spent bringing the product offering into a more marketable state. Fair enough.

My biggest concern always is how many of these days/weeks/months/years the customer is going to be expected to wait for the anticipated and perhaps inevitable improvements to occur. 3G is new in historical terms, but now after well over a year of being in commercial operation it should not be improper to anticipate and expect the service improvements to occur.

Not too early [anymore] for the spotlight to be shone in the right direction - if indeed that is the case.