Recalling that I have posted here at trewmte.blogspot and cellsiteanalysis.blospot over the years was to assist interpretation of data and testing for cell site anslysis and elements that can be used when conducting investigations, I have posted below a few of the weblinks to help this discussion along.
https://trewmte.blogspot.com/2014/07/csa-site-survey-method3mobility-models.html
http://trewmte.blogspot.com/2009/08/cell-site-analysis-csa-images-part-2.html
http://trewmte.blogspot.com/2008/11/mobile-phones-and-fringe-coverage.html
http://cellsiteanalysis.blogspot.com/
https://www.dropbox.com/s/g912o5dji9wkxfk/3G%20Networks%20position%20techniques.pdf
It is noteworthy the ITU in 2017 published a series of documents regarding call details record (CDR) and specified network data that could be captured in CDRs to assist a wide range of tasks to comprehend mobile phone movement caused by migration to determining trip travel and destination. These studies were conduct in Liberia, Sierra Leone and Republic of Guinea:
Liberia CDR reallocation D012A0000C93301PDFE.pdf
CDR Sierra Leone D012A0000CA3301PDFE.pdf
CDR Republic of Guinea D012A0000D03301PDFE.pdf
The reports identify how to obtain, collate, display overlay geodata/mapping and interpolation of the format specification that I rather think is highly useful to CSA investigations. The ITU source highlights CDRs capturing association with PoI, Trip Segmentation, Trajectory and Stay Points etc. I am simplifying in my summary what is undoubtedly more detailed discussion in these reports to show that 'time' and ‘location’ will be highly relevant.
CSA has not been without the knowledge regarding peak-time call traffic, density of call traffic, tracking etc and these are used in call analysis and CSA. In these reports though the defining stay points captured in the call records add useful evidence such as travel, location, co-location (if relevant), association (if relevant), landmarks, so on and so forth.
Consideration of trip segmentation in the report states ""Trip segmentation: Extract stay points from anonymized CDR data, and divide move/stay segments. Figure 7.4 explains how stay points are extracted by applying parameters and thresholds to CDR data." In this regard the threshold parameters for stay points are specified as 'Minimum Time Duration 15 Minutes' and 'Maximum Distance 300 Meter'. To assist further here is a useful image with data from the ITU Liberia report:
To extrapolate such detail require Trip segmentation, Stay point reallocation, Route interpolation, Grid-based aggregation and Visualization and so on. To dig into the detail to assist interpretation:
"Stay point reallocation: Reallocate stay points (Trip OD) to surrounding points of interest (POIs) with a certain probability and fil gap between stay/move segments. POIs are regarded as surrounding a certain cell tower if they are closer to the cell tower location than to the others (Voronoi tessellation). The reallocation is necessary because CDR location data is based on cell tower location, which means that all users in the same area have the same location. Reallocation can make the distributing of people more realistic or likely because POIs can be considered places where people are likely to stay or visit, such as shopping areas, residential houses, villages, and to which people are reassigned rather than concentrating on cell tower locations. A new dataset of POIs was constructed for this process by collecting data from the distribution of buildings from open access Internet data (see Appendix 2). Figure 7.5 shows how POIs are distributed in a city. Areas in blue indicate building POIs with extracted stay points, where location information originally based on antenna location, are reallocated."
Lastly, the reports published in 2017 discussed relevance to 2G, 3G and 4G.
No comments:
Post a Comment