Analytek - Mobile Phone Testing

Mobile battery testing software and services

Consultancy

Consultancy to identify an obscure mobile performance problem

The customer is a mobile manufacturer with a new model of mobile just having achieved type approval.

While the mobile passed the RF tests at type approval with flying colours, subjective performance of the mobile in a "real world" network situation suggested that a problem existed. Field testing often revealed unexpected call drops in difficult reception areas and early users reported that the product seemed to have inferior performance when compared to competitors' products. Despite this, attempts to reproduce the problem in the laboratory had failed to reveal any problems.

Problems of this nature can be difficult to track down, especially given that the test equipment with which a manufaturer is armed tends to be geared to reproducing the tests performed during type approval. One of the weaknesses of this testing is that specific aspects of the mobile's performance are targeted and tested in isolation. In a live network situation, all aspects of the mobile's performance must comply simultaneously. This distinction often allows problems to pass unnoticed through type approval only to be picked up when a mobile is used in a live network or in field testing.

After an initial fact-finding with the development engineers, Analytek were able to develop bespoke test software supported by on-site consultancy. Based on some of the methods used in type approval RF testing, but combining several measurements of the mobile's performance, the testing employed several methods of degrading the signal received by the mobile to simulate the conditions which would be present in a live network environment.

It was found in this case that by combining two of the RF measurements normally carried out in isolation, and performing the test over an extended period of time rather than the few minutes duration of most type approval test cases, a failure mode of the mobile could be provoked which would not have been possible to identify with conventional testing.

Since this failure mode had been identified with a test instrument rather than during field testing, the failure could be reproduced at will, with several different revisions of the mobile hardware and software. Furthermore, the extensive logging capabilities of the protocol testing platform could be employed to target the problem. With field testing such logging is simply not available.

By repeating this test and analysing traces from the mobile and test set, it was possible to pinpoint the faulty subsystem within the mobile and to provide information that allowed the development team to identify and remedy the problem.