It is six years since the stated case of R v Calland should have been a watershed moment for cell site analysis.
The court went into detail as to what constituted expert witness testimony. Whereas previously presentations showing the route of travel using cell site data might not seem likely to fall into expert witness territory, R v Calland turned that on its head. It quite clearly stated that as the jury is being asked to judge this information in deciding guilt and this requires opinion, it falls directly into the realms of expert witness testimony, and the FSR Code overtly describes Cell Site Analysis as expert evidence.
But in the years – and cases – that have followed, there has been continued reliance in court on police analyst product rather than that of an Expert. One would have expected a sharp uptick in challenges to such evidence in the wake of Calland; and yet it is only now we are seeing a groundswell of ‘offensive’ activity from criminal defence teams.
The role of cell site Experts is to serve the criminal justice system: their job is to help both jury and judge understand technical evidence, always acting impartially while presenting the facts and offering expert opinion. They can be instructed by the police/prosecution or by the defence but they are always working for the court. At CCL, we maintain that ‘no sides’ stance, marketing our services evenly. That said, based on instructions we’ve received over the past 18 months, defence teams are increasingly alive to the potential for challenge.
So, it seems like the perfect time for a cell site ‘download’, sharing our knowhow, and covering both cell site essentials and developments on the ground in a digestible Q&A format. The content comes from one of the UK’s largest – and one of the most highly regarded - cell site analysis teams with Expert status. They are recognised as leaders in the area of opinion evidence, with a number of peer-reviewed academic papers published and valuable contributions made to the Forensic Science Regulator's Code of Practice. What you are about to read is much like the evidence they give – clear, objective, and robust.
Data stored on mobile phones is a common source of investigative and evidential material, but there may be further useful information available such as Call Data Records (CDRs), held by network operators for billing purposes. This information is potentially available to be used within criminal investigations. Historic Cell Site Analysis involves the analysis of these records, in conjunction with other information such as survey and geographic data, to determine areas where a mobile device may have been at the time of call activity recorded by the network. This is also referred to as ‘Geo Location’ with the FSR code. The analysis may enable investigative insights as to the area in which the device was (and it is always an area) or potential movement. If used at court, an evaluative assessment of whether the data might (or might not) be expected given the positions taken by the Prosecution and/ or Defence can also be made.
The concepts of cell site analysis are relatively simple:
i) When a mobile device interacts with a telecommunication network, a record of that interaction may be retained which can include the cell site (mast location) and the specific Cell (unique to a specific area) used.
ii) These records may be available, with appropriate legal authority, to organisations within Law Enforcement and the wider criminal justice system.
iii) Inference can be made as to the area in which a device was at the time of the activity, and this assessment may be used to better inform an investigation or aid the jury in coming to their view on the guilt or otherwise of the accused.
iv) Where the records of multiple devices are available inference can also be made as to whether those devices may or may not have been together or in contact with each other.
A common misconception is that cell site analysis can pinpoint a device to a location; rather cell site analysis can only provide an insight to the area in which the device was. A more dangerous misconception is that cell site analysis is a simple, factual process and that it is safe to just present maps and tables from the call data to a jury for them to make their minds up on what it all means. There is uncertainty throughout the process; this includes initial generation of the call data on which the analysis relies (which is designed for billing, not forensic purposes) and through each subsequent process within the analysis (survey measurements in particular). This means that the execution of cell site analysis can be anything but simple, and it always results in opinion.
Unfortunately, unsafe practice is widespread in the field, for example presenting ‘factual’ maps and tables to the jury. Relying on 2D maps to draw safe inferences from the data is only safe in the broadest of questions (e.g. movement between cities) – and even then, can be problematic if it relies on GPRS data, which cannot safely be taken at face value. It isn’t just ‘facts and common sense’, there can be nuance. These issues concerning reliability within telecommunication evidence came to a head in R v Calland, in which there was a ruling by the Court of Appeal. One of the concerns raised by the Defence was that [The prosecution had] “…unhelpfully conflated the location of a particular mast with the question of what safe inferences can be drawn about the telephone using it”; this objection was upheld.
The FSR Statutory Code recognises the risk that poor practice in cell site analysis presents to the Criminal Justice System, and it defines acceptable (and unacceptable) approaches. CCL complies with the statutory FSR Code.
Mobile forensic experts specialise in recovering deleted data from devices. But from a cell site analysis perspective, the networks don’t keep content (e.g. a message) so it makes no difference if it’s deleted on the phone or not, we won’t see it either way. Cell site analysis isn’t concerned with content; it’s concerned with call data surrounding the content that is held by the network. The phone doesn’t even need to have been recovered for an analysis to take place, and a user can’t delete their Call Data Records (CDRs) even if they wanted to. Every UK network stores CDRs for up to 12 months, so as long as the request from Law Enforcement to the mobile phone company is within this period, cell site analysis may be possible.
There are three main ways in which cell site analysis can provide insight at court:
i) The activity of a phone in a given period – for example, whether the call data might be expected if it were at the scene or alibi locations.
ii) Aiding in an attribution assessment. If there is an alleged (‘dirty’) phone, we can compare that data with that which we might expect if it was used by the suspect. This can be done by comparing the ‘dirty’ phone data against known movements (via attributed, or ‘clean’ phone data, ANPR, observations etc.) and areas of common usage (compared to, for example, known home and work addresses).
iii) Aiding an attribution assessment of a vehicle used in an offence where the movements of suspect mobile phones mirror the ANPR captures of e.g. a get-away car before or after an offence. This analysis is vital in cases where the suspects turn off their phones whilst actually carrying out their offences.
Cell site analysis is particularly useful in cases where understanding someone's movements is crucial, such as in criminal investigations or family law cases.
A given phone number is associated with the SIM card, not the handset, but there is also a unique identifier for the handset – the IMEI (the International Mobile Equipment Identifier). This is often recorded in CDRs and, if the same IMEI occurs in the CDRs for more than one phone, it is a relatively trivial matter to detect this. This can be very useful as part of a wider attribution assessment, which can include other non-cell site factors such as numbers with which different phones (for example, an attributed contract phone and the alleged phone) have mutually been in contact with.
i) We never know where the phone was, but we may know the Cell ID to which it was connected at a certain time. Cells serve areas, and we can’t get away from that: no matter how well the industry is able to develop technical methods, it will remain very, very rare to be able to pinpoint the exact location of the phone.
ii) The CDRs we consider weren’t designed to help us work out where a phone was, they were intended for billing purposes. This means that if we take them at face value they can mislead; this is particularly pertinent to analysis of ‘data’ events (otherwise known as ‘GPRS’) in which, amongst other issues, the indicated time of a connection to a given cell may not be true.
iii) Estimating the area over which a cell serves can also pose a challenge, there’s a lot more to it than just plotting where the mast is. No two cell masts have exactly the same range, and the ranges of cells can vary enormously from a hundred metres or less up to tens of kilometres. Furthermore, the areas that cells serve can also change with time and this is another uncertainty we must accommodate in our analysis. All is not lost, though. We are often able to make an assessment of changes, and indeed have published research on this problem.
Additional information concerning the cell used for a given call event may be detailed in the Call Data Records. This detail can include the cell site location, azimuth etc. which can be used as a preliminary estimate of the anticipated service area (i.e. the cell is expected to serve from the mast location and into a wider area in the approximate direction of the azimuth). The estimate can be refined by considering survey data, terrain profiles, expected or known presence/ absence of other cell sites etc.
Survey data can be very useful when doing this, but it is not perfect. At CCL we have state-of-the-art equipment, and our validated survey methods include the ability to monitor multiple networks and technologies simultaneously. We store this data in a database, currently running at over 800 million measurements; it helps us estimate the total service area of a cell and also whether there may have been changes to the network between the times of the calls in the CDR and the times of any targeted surveys tasked to aid in the specific case we are working on. This can be very powerful when informing opinion as we have a breadth and depth of ground-truth data built up over many years.
Within section 110 (“Cell site analysis for geolocation”) of the Statutory FSR Code, it states:
“110.9.10 The location of the device is not in the CDRs, but the area in which it was operating may be inferred from the detail within the records. Inference is synonymous with opinion, and therefore should be given by a practitioner who is competent to provide evidence of opinion (i.e. an expert). If the purpose of the analysis is to assess where the phone was, this will require the practitioner reporting the finding to be giving an expert opinion.
110.9.11 Expressing opinions is the role of the expert witness; this includes providing evaluative evidence. Practitioners interpreting results shall have been assessed and deemed competent in interpretation and opinions”.
All reports containing opinion from CCL are issued by a competent Expert who complies with the rules, regulations and standards concerning expert evidence for England and Wales.
The quotes above from the Code mention ‘Evaluative’ evidence. This is when the probability of the evidence (i.e. the CDRs) given the position taken by prosecution is compared to that of the defence. This is in contrast to ‘investigative’ opinion, which provides speculative explanations for the data. Confusing the two can cause serious issues in opinion evidence at court, potentially resulting in ‘the prosecutor’s fallacy’ (or ‘Transposed Conditional’) which can be grounds for appeal. This is another reason why just presenting ‘factual’ prosecution-selected maps and tables to explain why the suspect did it is an unsound approach to evidential cell site analysis.
The FSR Statutory Code defines best practice; it is the de facto standard for cell site analysis in England and Wales. There is no obligation to comply with it until October 2025, but if practitioners don’t comply, their work may end up being considered to be of a lower standard. Criminal Practice Directions (related to Criminal Procedure Rules) includes use of expert evidence; CrimPD applies now. At the core of the Codes is the requirement for a practitioner to be able to demonstrate the validity of the methods they use and their own competence.
This means that some current practice does not comply with the Code and can be challenged. For example:
i) The production of maps and tables and presentation to juries without overt opinion on the meaning of them in the context of a case.
ii) Impressive CVs with reports essentially saying “I’ve been doing this for ages, it matches because I say it matches” without auditable documentation of validation, competency, processes, information security etc.
CCL’s Standard Operating Procedures are based on, and comply with, good industry practice as defined in:
Formulation of ‘opinion’ by CCL Experts follows the Case Assessment and Interpretation Model as extensively used in traditional forensic science (such as DNA, glass, fibres etc.). The processes adopted by CCL are those considered ‘Best Practice’ by the European Network of Forensic Science Institutes (ENFSI), whose membership contains governmental laboratories in 39 countries (https://enfsi.eu/wp-content/uploads/2016/09/m1_guideline.pdf).
Experts may share a title, but they don’t always share the same approach, mindset or background. Technical output can vary, as can performance on the witness stand. So how do prosecution or defence teams ensure they really do get the expertise they need? Here’s one very useful pointer.
One of the key responsibilities of an expert in any field of forensics is that they remain balanced and impartial and are aware of and avoid the potential for bias in their analysis.
In the 1970s there were several miscarriages of justice, such as R v Preece (1972), R v Ward (1974) and R v ‘The Birmingham Six’ (1975). To address this issue, the Case Assessment and Interpretation model was created, focusing on areas such as impartiality, potential bias, and the understanding and expressing of uncertainty.
The Association of Forensic Science Practitioners (AFSP), whose members include the Governmental labs of Scotland, Eire, Northern Ireland, and the largest private sector ‘traditional’ forensic labs in the UK, have adopted this model to avoid potential further miscarriages of justice. It is also endorsed by the European Network of Forensic Science Institutes (ENFSI), the Royal Statistical Society, and the Chartered Society of Forensic Sciences. This model is, however, yet to be commonly adopted by Experts in the area of cell site analysis.
All of CCL’s cell site experts have been trained in, and employ this model, to ensure that the evidence provided is robust and as balanced as possible. This not only reduces the risk of miscarriages of justice, but also the possibility of their evidence being deemed as unsafe for presentation at court. On the flipside, the grounding in this model also makes CCL Experts alive to the opportunity to challenge their peers on behalf of the defence.
We can but must preface things by stressing that we don’t ‘take sides’. We serve the criminal justice system by providing Expert cell site witnesses both to the prosecution and defence. Acting impartially, we present the facts and offer expert opinion
The following case study is instructive. In spring 2024, CCL was brought in to assist in a complex defence case involving the supply of class A controlled drugs, with cell site data key to the investigation. The prosecution alleged the defendant was involved in the offence due to a mobile phone (not accepted by the defendant) connecting to cell site near to the defendant’s home address, and the same phone having been in contact with people known to the defendant.
A Police Investigator (not a cell site expert) produced several cell site exhibits and provided statements regarding the call data.
One of our cell site Experts reviewed the prosecution material, raising several issues including:
Our Expert conducted their own analysis, with their report detailing the limitations, uncertainties and conflicts found in the prosecution material. After receiving the Expert report and challenging the prosecution, the defence team later reported that the prosecution submitted no evidence, and the case was withdrawn.
This is as good an illustration as any of why all of us working in the criminal justice system need to stay alive to the issues around cell site analysis.
It’s fair to say that the introduction of the FSR Code has presented a challenge to the way that many organisations (LEAs, Companies and lone practitioners) operate. A discussion of the statutory FSR Code and its effect on current practice (mainly ‘factual’ analyst product presented straight to the jury) can be found here.
Technology changes, understanding of artefacts changes, approaches to inference can be affected by ruling in other areas of forensic science. Experts must stay on top of developments, so R&D is critical. CCL has an active research program addressing every aspect of cell site analysis, including assessing uncertainties in the call data, development of novel methods to analyse it, and how to formulate sound opinion. Some elements of our research have been published in the academically peer-reviewed Q1 scientific journals (“Science & Justice” and “Digital Investigation”):
i) “Historic Cell Site Analysis – Overview of Principles and Survey Methodologies”, https://doi.org/10.1016/j.diin.2011.10.002
ii) “Cell site analysis: Roles and interpretation”, https://doi.org/10.1016/j.scijus.2019.06.005
iii) “Opinion evidence in Cell Site Analysis”, Awarded a PW Allen commendation from the Chartered Society of Forensic Sciences as one of the “most meritorious” in any field of forensic science that year based on a range of factors, including the quality of the science. https://doi.org/10.1016/j.scijus.2020.02.002
iv) “Cell site analysis; Use and Reliability of Survey Methods”, https://doi.org/10.1016/j.fsidi.2021.301222
v) “Cell site analysis: Changes to networks with time”, https://doi.org/10.1016/j.scijus.2022.04.001
vi) Elsevier Major Reference Work (MRW) - “Encyclopaedia of Forensic Science (3rd Edition) ISBN 978-0-12-823678-9. Chapter: “Forensic Inference from telecommunication Records”
Collectively, these papers form much of the academically peer-reviewed foundation for the field. The experts who wrote them are available for consultation by both prosecution and defence teams.
Aside from all court-going experts being CAI trained, competency tested, using validated methods and complying with the FSR statutory Code, there is something else we point to in highlighting the value of our team. Because it is a team, and while experts are individually appointed, this is a group of professionals who share knowledge and experience: together it aggregates some 300 years of deep and diverse knowhow of radio, networks, forensic science, standards and R&D, pooled from careers spanning telecommunications, military, law enforcement, electronics, engineering and research. Each expert then has their own track record in the field, ranging from five to twenty years’ specialisation. Quite simply, this is an unrivalled talent pool - there is no team in Europe more capable or more proven as cell site experts. And in Dr Matt Tart, we have the world’s first and only holder of a doctorate specifically in forensic inference from telecommunication records i.e. Cell Site Analysis.
To discuss any aspect of cell site analysis or to instruct one of our experts, please contact Brian Edwards – brian.edwards@cclsolutionsgroup.com – in the first instance.
Our experts are on hand to learn about your organisation and suggest the best approach to meet your needs. Contact an expert today.
Get in touch