Analysis of GSM networks performance with data mining

All institutions that use information and communication technologies, save the data which is produced with computer systems and electronics devices to databases of their institutions. Data mining is accessing and using the meaningful information job from the very large amount of information stored databases which will allow us to predictions about the future in line with our purposes. Parallel to developments in information and communication technologies, software technologies that used in these areas are in continuous improvement. The new favorite of people has been ?mobility? which means to be independent of time and place and freedom of movement. Recently, GSM Networks subscriptions as the number of population of the country, mobile communication devices at everybody, lots of person?s two subscriptions usage, and using mobile communications not only for communication but also for video communication and internet shows that mobile communications are very important for our life. Variety of this usage area and intensive usage made very big amount of data in mobile communication. Thus, using data mining in mobile communication became inevitable for company owners and managers. In this thesis, a program is developed with using MS Visual Stdio C#.NET 2008 programming language which creates database by taking three separate GSM networks? signal data from mobile phone. The data which is stored from mobile phones is made available for data mining and then is analyzed with Frequency Analysis and Two Step Cluster methods and results are declared. These results showed that A Network has the most powerful range of signal quality and the most stable measurement. Then B Network follows the A Network. C Network has the lowest performance for the signal quality range. Also other findings are declared and commented.