Visualization of a radio signal coming from a mobile phone in a data filled scene.

How my research into the accuracy of follow-on GPRS CDRs, published in Science & Justice, can benefit criminal investigations

Isabel Duncan graduated with a distinction in Crime and Forensic Science MSc from UCL. She now works for Forensic Analytics as a Forensic RF Technician.

Technology is constantly evolving. In the case of mobile phones, this evolution has seen a shift away from phone calls and text messages to messaging services and apps such as WhatsApp, Facebook Messenger and Instagram.

Included in this shift are those involved in criminal activities – indeed, it was recently estimated that more than 80% of crimes now have a digital evidence component.

Call Detail Records (CDRs) provide timestamped intelligence relating to the geographical area within which a device could have been during a crime-related event. The analysis of these records is a discipline known as cell site analysis.

Each time a device is used to make or receive a phone call, text message or uses mobile data, a CDR is created within the billing record for that mobile subscription. Follow-on GPRS CDRs are currently considered inferior to voice or text CDRs due to uncertainties regarding the correspondence between the timestamp and the Cell ID presented within the CDRs.

So far, very little published research has been undertaken to test these uncertainties. With a lack of empirical research, there is a danger that incorrect interpretations are formed or important data is disregarded from an investigation due to concerns for its reliability.

This research gap lends itself nicely to become a suitable research subject for my dissertation whilst studying for an MSc in Crime and Forensic Science at UCL. I developed a detailed investigation into the accuracy of follow-on GPRS and mobile data CDRs by conducting connected mode radio frequency (RF) surveys while simultaneously producing follow-on GPRS/mobile data CDRs in a travelling vehicle.

How could I achieve this? I was helped in my task thanks to the support of Forensic Analytics, principally one of its founders Joseph Hoy, and Training Director Andrew Fahey.

With equipment and support provided by the company, one of the UK’s leading providers of innovative software solutions in the fight against crime, I focused on investigating the uncertainties associated with follow-on GPRS CDRs.

These were conducted using two Samsung Galaxy S9 phones equipped with the RF surveying software ‘Nemo Handy’ so 3G and 4G technologies for each network could be tested simultaneously. The data was then processed and analysed using Forensic Analytics’ CSAS and CSAS RF Survey.

The full text of the investigation has recently been published in the journal Science and Justice, and you can access it here to take a look at our findings. It certainly struck the right note – my work saw me jointly awarded the UCL Locard-White prize for best MSc Crime and Forensic Science Dissertation 2020.