Design And Implementation Of Diagnostic Tool To Optimization Of Base Station Based On Crowdsensing

With the rapid development of Mobile Internet, the applications of 3G, 4G are broadening constantly, the number of mobile phone is increasing year by year, so that relevant base station (BS) construction plays an important role in this field. However, the BS often produces some switch problems due to geographical location and unreasonable parameter setting, such as ping-pong effect and islanding effect, directly affect quality of call and increase the power consumption. At the same time, the diagnosis to ping-pong effect and islanding effect faces some difficulties which need to consume much people and money, other situation, the diagnosis are not complete. In order to solve these problems, in consideration of the background that large number of android mobile phone, this paper designs and implements a diagnostic tool for ping-pong effect and islanding effect based on C/S architecture.Works in research and development are on following aspects:(1) In this paper, it analyses the related technologies about Android system, relevant knowledge in switch of BS, reasons of ping-pong effect and islanding effect, decision tree model and ID3 algorithm which is used to research the diagnosis of BS optimization.(2) According to the characteristic of the ping-pong effect and the islanding effect, this thesis defines the requirement for this tool and proposes the structure, framework of the diagnostic tool that based on crowd sensing with flow chart, structure chart, and design the functions what the tool should have. This tool could be divided into client module and server module. The client is divided into:data collection module, data filter module, data analysis module; Server is divided into:data classifier, training module and diagnosis module.(3) The ping-pong effect, islanding effect model and the diagnostic methods are presented, then, this paper implements the key technology for each module. It makes some pre-processing and processing to the data received by server, uses corresponding ID3 algorithm and diagnostic manners to these data for analysis and diagnosis to determine whether there are corresponding coverage problems of BS. Finally, according to the location and result of the diagnosis, some flags are marked dynamically on the map in real-time, so people can easily find blind spots, ping-pong effect location and islanding location. According to the analysis and comparison from the test result of the tool, the diagnostic tool based on crowd sensing which design in this paper can minimize the energy and data flow, makes full use of the characteristics of crowd sensing and collects the data accurately for each latitude and longitude, makes judgments accurately. The means used in this paper are feasible and effective, with good usability.Firstly, the tool in this thesis uses crowd sensing and large number of Android terminal to make the diagnosis more intelligent, it can collect more data about BS make the diagnoses accurate. Secondly, ID3 algorithm and decision tree model are applied to the diagnoses of ping-pong effect for BS. It is not only complete the diagnoses for BS, but also give data support to optimization for BS.