Research On Comprehensive Optimization Methodology For Mobile Communication Networks Based On User Behavior

The mobile arena has been expeirencing exponential growth in the past 20 years and the technology has matured--now the 3G network is developing into LTE,4G network. The feature and demand of present mobile communication network include: heterogeneous network environmentsin, innovation for network service, diversified network data and so on. In order to build a new generation mobile communication network characterized with heterogeneous, ubiquitous service, data diversity, we must take the present network as the basis and consider business integration as the goal. Under the support of National Natural Science Foundation project "Research on dynamic coacervate mobile network model with the Support of group properties" and National Science and Technology Significant and Special Project "Opening Research on Critical Technology of Broadband Mobile Business", this sbuject will mainly concern on network optimization work based on social attributes in current hybrid network and the further implementation of the construciton of LTE and 4G network. Through the analysis of mobile user behavior, we will discuss about network comprehensive optimization in mixed network construction which will promote the development of research on mobile communication networks. Unlike traditional network optimization methods, network comprehensive optimization consider not only how to improve the quality of network service but also how to improve users' feeling when they are accessed network--in other words, to provide more inovative services on the base to meet the basic communication needs. User behavior need to be taken into consider.The main contributions of the presentation are listed as follow:collect and analyze the data which reflects user behavior, then classify it to find the relationship between the data and user behavior; model the user behavior characteristic data to get the statistical nature of it, further study of the traditional network optimization methods including BER improvement, hand-off and synchronization problem etc; discuss the users' mobility, consumer model, interaction model etc to investigate the regularity of user behavior, which may be helpful to serve users better. The details are as below:(1) The moethod of investigation mobile user behavior includes:collection, perception and analysis of user behavior data; formation the direction and goal of user behavior analisis; join key performance indicators (KPI) and user behavior; to be userful complement and insprition of traditional optimization; construction of mobile network user behavior analysis system in order to serve user better; feedback user behavior indicators to improve the KPI.(2) Conllect the data relecting the user behavior for analysis and summary, then a realizable scheme is presented. These data is used in locationization and relevance anaysis to get user behavior statistical parameters for better network performance. Analyze the characteristics of the user experience data to improve and optimize the network and give a new method for studying the network model and user behavior.(3) Investigate and present new methold founded on user behavior for traditional network opitmiziation problems such as hand-off, the establishment of network synchronization, system BER, and a framwork for it is also presented.(4) Through the distribution of telephone traffic of base station to investigate user moblity behavior, especially for groups of mobile user, present methodology of user location predication which can be the base of network location service and improve users' feeling when they are accessed into network.(5) Model user communication behavior based on the incoming/outgoing call holding time and then use fuzz c-means clustering algorithm to classify every level in user pyramidal model. The method and conclusion can be used as the base of precision marketing for telecommunications industry.(6) Construct communications through user interaction social network, composed by the user and the contact link modeling and analysis, link stability prediction and forecasting methods are given.This presentation focuses on the mentioned six issues. Through the research on network topology, business classification, communication standard, functional entities, we descript the interactions between network behavior and network entities under several communication modes in detail.In this presentation, after a series of data collection, data analysis, data fusion, algorithm improvements, simulation comparison and case studies under this line of thought, the analysis conclusion based on user social behavior in the hybrid network has been acquired.Such study is one of the exploratory researches which inspired by subjects that need urgent solution to the construction and design optimization problems of hybrid network in mobile communication field. This presentation will advance the development of future mobile communications network technology, and make a breakthrough in the mobile communication system construction process. From such point of view, this paper possesses certain significance of theoretical prospect and practical application.