Performance

In-Field Performance

Network operators constantly monitor call drop rates and RAT selection and average dwell times for the population of devices active on their networks. These metrics can be used as acceptance criteria as well as for operational reasons.

Automated Log Analysis and Statistical Analysis

Determining whether or not a high rate of call drops is a device specific or network specific issue is relatively straight-forward for the operator. However, determining why there is a high call drop rate for the handset or device manufacturer can sometimes be very difficult. Sensitivity issues may not be the underlying reason, but instead subtle interactions with the local network infrastructure or more fundamental issues with security procedures or other protocol stack issues interacting with the macro or small cell network elements may be the cause.

Automatic analysis of device side logging to classify call drop failures, feeds into a statistical analysis process and then detailed specification knowledge and practical field experience are often essential to understand the true picture.

For accurate results this quickly becomes a Big Data problem and this is where we started our own Big Data journey !

Issue Recreation

Once the problem scenario is identified, it is usually desirable to recreate this under lab conditions to avoid expensive field trips. We can assist in modelling the relevant network conditions, then designing test cases to run under system simulation. Finally providing analysis of the results, proving where the issue lies and suggestions of any work-arounds possible.

Relevant Skills

Experience in designing automated log analysis functions.

Detailed specification and protocol behaviour knowledge to identify problem areas and do root-cause analysis.

Flexible test creation ability, use of programmable faders for recreation of network cell environments.

ALCHEMY WIRELESS LTD

(C) Alchemy Wireless Ltd, 2016