Many of the cases I have done included evidence presented incorrectly;
Any reason why were you called to give evidence on this subject? Is that because this is your specialist subject?
Larry,
You bring up a lot of good points. It seems to me that most of the things you mentioned were due to sloppy work. I had a conversation this week with someone that teaches call detail records analysis and he is not an advocate of drawing circles or pie shapes. As he suggested maybe we just need to concentrate on the location data as it pertains to the originating and terminating tower? Sometimes that might be all we need? If the law enforcement community, of which I am part of, is trying to get convictions based only on the location of a hand set then it’s a sad day. As good as this stuff is, it is only a tool.
I’m sure you are aware that the two major CDMA providers do retain enhanced location data although it is only kept for a short time. Do you have any comments on its accuracy?
I would submit that the engineering data you were referring to is available, but it is difficult or cost prohibitive to obtain.
You wrote The radius of towers is not provided by the carrier other than the "Design Radius" which is normally stated at the maximum range of coverage, i.e. 3 miles, 9 miles, etc
Are there providers or analysts determining this radius based solely on the maximum range of coverage? If we used a three sided mast as an example I would theorize each sector is at least capable of 120 degrees of coverage, but I do know that there are dead spots. Depending on the needs of the carrier the panel is adjusted to face towards the area the carrier has the greatest need. Horizontal beam widths, down angles, PN codes and other factors are set into the panel in an attempt to make the tower work efficiently. What is the “Design Radius?”
I’ll go ahead and concede that you probably have forgotten more about locating a handset based on historical call detail records than I will ever know. I’m not trying to pick a fight, but I would appreciate it if you could tell us where you received the training which qualifies you to testify in relation to a handsets historical location. I didn’t see your experience in this field listed.
Is your book in print yet? If not let us know when it hits the market.
Those are all fair questions. I received my training from an RF engineer / analyst that does a ton of these cases. I then supplmented that training with self study by reading books by William C. Y Lee, and others, as well as on line training through the Teracom Training Institute.
I have qualified and testified an a expert witness in this area on multiple occasions in NC and MO.
As to tremtes question as to this being a speciality area, is it one of my areas of expertise. I have also qualifed and testified as a computer forensics expert and cell phone forensics expert in multiple states and in federal court.
Remember that the threshold for expert testimony is that the expert have knowledge gained through training, education and or experience that is greater than the average person and that would assist the trier of fact to understand the evidence. I am paraphrasing, but that is the basic standard in the US.
More worrisome than the expert threshold is that persons are being allowed to testify as lay or fact witnesses about this technology. That is fine if all they will testify to is the locations of the towers and the towers that were used for calls. However, anything beyond that goes into expertise that many do not have.
You don't need to be an RF engineer to analyse this kind of data, but you should have a firm qrasp of the way the systems work, the different technology, (CDMA, GSM, etc.) and how systems handle cell calls.
I agree that while engineering data may be something you can get if you can get the cellular company to provide it, but it is not necessarily going to be that helpful in many cases due to the many factors that go into how towers are set up, changes in equipment, etc.
By design radius, let me give you an example. I recently got the cell tower location information for a carrier in a major US city. For each tower, it had a radius number in miles or meters.
If you plotted the towers using the provided radius, every tower would overlap the coverage of at least 10 other towers, sometimes more. All of the towers were listed at at least one and a half miles and some of the towers were listed at 9 miles radius in the city.
However, bear in mind that some analysts simply default to a number they were taught by someone, i.e. Towers are 1 mile or 3 miles and they plot the towers like that with no regard for other towers in the area.
Also, you do not always know if a tower is an umbrella tower, i.e. an Omni that is used to pick additional calls for an area when call volume exceeds that of a particular tower.
I am referrinf to several posts here so bear with me.
As far as being able to know which tower handled a call, that is in the call detail records. To say it is impossible to know that is a misunderstading I think.
As to WHY the call was handled by a particular tower is somewhat squishy if all you have is the call detail record without the RF plan from the carrier for that date and time, the trouble tickets and tower loading that would assist in determining that the towers in the area are all functioning and able to handle calls.
Also, if you get an a call that is handled by tower A and another call immediately after uses a tower many miles away, is it safe to assume that it was transferred via a backhaul due to tower loading? How can you determine that from a call detail record? Not with any definition.
I am not going to attempt to cover every scenario in a post, but my point is that there is a lot of wonkiness in the reports and analysis being done.
As far as the enhanced location data, you are right, it is only available for a short time, so I rarely see it. As to its accuracy, it is better than just knowing which tower was used.
Even e911 phase two calls that have location data will many times have a confidence number that gives a radius that is less than the legal requirement for that type of call. In an exigent case I reviewed, the radius was 400 meters, not the 150 to 300 feet it is supposed to be. However, for the purpose of that case, it was close enough.
I think my fundamental issue with the analysis being done in the US today is that it is regarded as a simple process that can be based on "default" values learned in a few hours of training or anecdotal informatino provided by others, rather than careful analysis of the records, the technology type and other factors. In one case where I testified, the analyst entered the call detail records into PenLink and it told him to use a default three mile radius for the towers. That was the extent of his analysis. And it was wrong.
And I agree that in many cases, all you really need is the tower locations, not the pie shapes. But excessive claims are being made about the accuracy of cell phone locations from historical call detail records that is not supportable. For instance, claiming that if two cell phones use the same tower, the persons must be together.
When someone is facing a penalty of death or life in prison, I think this kind of forensics should be considered very carefully as to it accuracy and the analysis performed.
As far as the 120 degrees for a three sector mast, you are correct, but it is better to know the beam width and the azimuth, which are not always obtained. When that data is not obtained, it is not correct to make assumptions.
A three sector mast is not always going to have antennas that cover the 120 degree sector area, but may in fact only cover 30, 60 or 90 degrees.
Anyway, that is probably way too much information in one post.
Larry
There is only one point I would agree with CSA and that is it is not a precise science. However, CSA can and has provided useful evidence in cases.
CSA involves many aspects and the physical configuration at the Mast (cell tower) is merely one aspect of recorded information amongst a huge range that is available eg
There is a vast distinction between what is available as opposed to what people will let you believe is only available. CSA is relevant and applicable on a case by case basis. Applying a wholesale approach to suggest every case is the same, even where similarities can appear between cases, is a fundamentally flawed approach.