Evidence is key to any court case and having access to a large pool of data that can be used as evidence is always advantageous.
In 2105, New Zealanders made 6.63 billion minutes of mobile calls and sent 12.1 billion text messages1. That is on average around 1500 minutes of calling and 2600 texts per person. Each of these activities has a time, location and sometime movement data associated with it and all of which, could potentially be used as evidence.
Overseas, cell phone call data is regularly used:
in domestic violence, murder, robbery, drug and rape cases,
to challenge an accuser’s or suspect’s credibility,
to corroborate or destroy alibis,
to track suspects’ movements and
to prove when and where the murder took place, and connect the victim to the suspect and accomplice
Cell phone data provides an avenue to independently corroborate or challenge other crucial evidence.
Prosecutors used David Westerfield’s cell phone records to track his movements in the days after Danielle van Dam disappeared from her San Diego home. Within 48 hours, Westerfield drove to the desert, the beach and back to the neighborhood where he and the Van Dams lived. He told police he was in the desert scouting for places to take his son camping. Westerfield was convicted of kidnapping and murdering the 7-year-old girl4.
Aaron Graham and Eric Jordan from Baltimore were convicted for six armed robberies in 2010 and 2011. Investigators used their cell phone location data for more than seven months of activity to link them to the crimes. There were no eyewitnesses tying Graham and Jordan to several of the robberies, so their prosecution hinged on the cell phone data5.
Law enforcement used records from Davis’s cellphone carrier, to establish where he was during a crime spree in Florida in 2010. Investigators got call records for 5,803 calls and 11,606 location data points, an average of around one location data point every five and a half minutes for 67 days. That allowed them to put him in the area of robberies. He was convicted and sentenced to almost 162 years in prison6,7.
Rebecca Cleland was accused of hiring her two cousins to kill her husband. Cell phone records showed 11 calls between Cleland and her cousins in the hours leading up to the murder, including one last call 10 minutes before the shooting. Cell phone records placed her cousin Alvaro Quezada about a block from the murder scene, despite his claim that he was at a restaurant 20 miles away. All three were convicted of first-degree murder4.
Cell phone call data has been successfully used in UK as well8
On 22nd August 2007, Rhys Jones, was shot dead in Croxteth, Liverpool, an innocent victim of a feud between two rival gangs. Sean Mercer was convicted of murder and six others were convicted of lesser charges. Communications data was used to attribute telephones to each of the offenders, demonstrate association at key times and place individuals at specific locations. It also showed that the telephones of the key offenders were in the area some twenty minutes after the murder – helping to establish that Mercer and other convicted associates attended business premises in order to burn the gunman’s clothing and douse him in petrol to remove firearms discharge residue.
In all these examples cell phone location data was used to place suspects in the area of crime and to corroborate or challenge other evidence, which led to successful prosecution.
Data Retention Requirements
All of this is possible only if you can get access to the call data. Data retention requirements vary from country to country.
Table 1: Data retention period in Australia, USA, UK and New Zealand
|Australia||Minimum 2 years|
|USA||Minimum 1 year. Some operators (AT&T) has data from 2008.|
|UK||Minimum 1 year|
UK has a Voluntary Code of Practice on the retention of communications data since 2003 (updated in 2007 and 2009). In Australia, the Telecommunications (Interception and Access) Amendment (Data Retention) Act 2015 made it mandatory for providers to retain data for a minimum two-year period9.
To our knowledge there are no requirements on the telecommunication providers to retain call data in New Zealand. How long the data is kept varies, and it is determined by the operator’s business practices. Some level of detail is kept for few months, however, the detailed data that allowed the type of tracking we saw in US may be kept only for days. Therefore, the key is to put an order to protect the data while the necessary paper work is being done.
Some Real New Zealand Examples
In this section, we use modified cell phone data to show examples of insights that can be gained. The insights will vary from case to case, however, these examples should highlight what is possible.
Disproving Location Claim
Person A claims that the incident happened at his/her house (the House) and he/she never left the House on the day of the incidence. We show how this claim can be disproved even when the cell sites that were involved in the calls were from the general area of the House. Generally, it is hard to make such conclusions when all the cell sites involved are from the general area, however, in this case detailed analysis based measurement data show that he/she was not in the House the whole day.
Person A had the following call records on the day.
Table 2: Call data of Person A.
|Call No||Time||Call Duration||Serving cell at the start of the call||Serving cell ID||Serving cell at the end of the call||Serving cell ID|
|1||7:12:47||2mins 50sec||Cell Site 1||2826||Cell Site 3||7124|
|2||9:07:12||11mins 13sec||Cell Site 1||2826||Cell Site 4||5265|
|3||10:22:34||33sec||Cell Site 1||2826||Cell Site 5||2870|
|4||10:54:24||45sec||Cell Site 1||2826||Cell Site 1||2826|
|5||12:10:18||5sec||Cell Site 2||2276||Cell Site 2||2276|
|6||12:25:08||6sec||Cell Site 2||2276||Cell Site 2||2276|
To establish whether the claim was true or not we consider the question “whether call events 1-6 can happen if the person was in the House?”. To answer this at a technical level we consider two aspects:
What is the probability of cell sites in Table 2 serving the House and
What is the probability of handover (where cell sites at the beginning and end of the call are different) happening if Person A was in the House.
To carry out the analysis we need to consider other information that are relevant to answering these questions.
cell site location data (obtained from Ministry of Business Innovation and Employment (MBIE) database) in relation to the House,
signal strength measurement data at the House and
propagation and statistical analysis.
Figure 1 shows the locations of the cell sites that were involved in carrying the calls to Person A in relation to the House.
Figure 1: Cell site locations
From measurements taken around the house we worked out the probability of each cell site serving the House.
Table 3: Probability of each cell site serving the House
|Cell Site||Probability of serving the House|
|Cell Site 1||5%|
|Cell Site 2||93.75%|
|Cell Site 3||1%|
|Cell Site 4||0.25%|
|Cell Site 5||0.0%|
We can conclude with certainty that Person A was not in the House when Call 3 was carried by Cell Site 5.
The probability of Calls 1 and 2 taking place in the House is very low (less than 0.025%). One should also consider the likelihood of two very low probability events happening consecutively, which may lead to the conclusion that the person was not in the House during that period.
It is likely that Person A was in the House during calls 5 and 6, but it does not conclusively prove that Person A was in the House.
Only limited data was available for the calls presented in Table 2. If more details were available, one could gain greater insights. For example, conclusively prove that Calls 1 and 2 did not take place in the House.
Direction of movement
Sometimes call records can conclusively prove movement. For this we look at Call 3.
For the call to be handed over from Cell Site 1 to Cell Site 5 the person had to move to the red highlighted area.
Figure 2: Movement of Person B during Call 5.
Table 2: Phone records of the accused
The distance between Cell Site 1 and Cell Site 5 is about 2km and the call lasted 33 seconds. From this we can make the following conclusions:
The person was in the blue highlighted area at the beginning of the call
The person was travelling at speeds around 100 km/hr. to cover such a distance
The person was travelling southwest.
This information can be used to either corroborate or challenge other evidence. For example, if the suspect claimed that he/she was near the blue highlighted area and travelled north and has an alibi, the above information can be used to challenge the claim.
Large poll of cell phone call data is available that can be used as evidence,
Overseas, cell phone call data is regularly used in court cases and sometimes as the primary evidence,
Cell phone data can independently corroborate or challenge other crucial evidence,
Key to accessing detailed call data is to get in early.