پيش ‏بيني خرابي BTSهاي شبكه موبايل قبل از ورود به سايت با استفاده از تكنيك‏هاي داده‌كاوي

Prediction of the faults in BTS mobile networks before entering to the site by the use of data mining techniques

Nowadays, one of the most important tools for communication is mobile network, it has a significant role in our life and people use the mobile phone regularly. In this research, we want to predict some faults in a BTS by using data processing. Predicting and recognizing the sorts of faults in seven groups consist of : ( transition , neighborhood , hardware , lack of traffic channel , lack of signaling channel , lack of traffic and signaling channel , similarity of frequency) and correct group were classified. The main aim of this project was related to propose the suitable frame for predicting the type of possible faults for BTS in mobile networks. In this method, the combination of decision tree algorithms, KNN, SVM were used. Firstly, chaos matric for every algorithm of decision tree, KNN, SVM was obtained. In the next step the value of precision in every algorithm for every cluster was determined. After that, the maximum precision in every cluster by the specific algorithm was detected. By the use of this technique, we can use the best algorithm for predicting every cluster. This method increase the rate of total precision. The combination method is more practical and proficient in comparison to other methods. In this method the rate of classifying accuracy, precision, recalling were better than other methods such as basic classification and other classification methods.

Nowadays, one of the most important tools for communication is mobile network, it has a significant role in our life and people use the mobile phone regularly. In this research, we want to predict some faults in a BTS by using data processing. Predicting and recognizing the sorts of faults in seven groups consist of : ( transition , neighborhood , hardware , lack of traffic channel , lack of signaling channel , lack of traffic and signaling channel , similarity of frequency) and correct group were classified. The main aim of this project was related to propose the suitable frame for predicting the type of possible faults for BTS in mobile networks. In this method, the combination of decision tree algorithms, KNN, SVM were used. Firstly, chaos matric for every algorithm of decision tree, KNN, SVM was obtained. In the next step the value of precision in every algorithm for every cluster was determined. After that, the maximum precision in every cluster by the specific algorithm was detected. By the use of this technique, we can use the best algorithm for predicting every cluster. This method increase the rate of total precision. The combination method is more practical and proficient in comparison to other methods. In this method the rate of classifying accuracy, precision, recalling were better than other methods such as basic classification and other classification methods.