Historically, Mobile Operators have been spending a lot of time and money on traditional drive testing to monitor and optimize the quality of their mobile networks. Using crowdsourced data is now changing that. Today, BI Nordic, with 20+ years of experience in the telecom industry, key competence within data analytics and with cooperation with various mobile date providers, is presenting a new way of thinking and a new way of working within this area. To be able to give as accurate and statistically confident results as possible, high volume of data is of key importance.

The crowdsourced data is being refined and cleaned in the BI Nordic solution. At this stage, more than 90% of the crowdsourced data is being excluded by not fulfilling the predefined. The remaining, high-valuable data is then being carefully analyzed using novel Machine Learning techniques. The outcome will be a list of network issues as well as proposed solutions to those issues.

What are the advantages of using crowdsourced data compared to traditional drive testing data?

Time- and Cost-saving

If using crowdsourced data, the operators would be able to notably cut down on the cost for their currently expensive investments of drive testing- as well as OSS-tools. The costs for well-educated drive testing teams collecting data in the dedicated drive testing vehicles could also be reduced

High correlation to end user behavior

Crowdsourced data is collected when/where/how the user is using the network. This means that the collected data can reveal when during the day the users are mostly active on the network, what places/locations the users are visiting, what technologies and back-bone vendors are used.

When can crowdsourced data in combination with BI Nordic analytics tool be used?

Network optimization

Makes is possible to identify network issues such as coverage holes, interference, overshooting- or backloob- issues.

Network rollout

When commissioning new cells or technologies, verify coverage or parameter settings, expanding network (hot spot clusters, load density etc.), identifying busy hours etc.


Benchmarking and trend analysis of already deployed networks, features and/or operators. Cell density per operation or region etc.