Optimizing System Performance inHigh Density Wi-Fi Networks

碩士 === 國立臺灣大學 === 電信工程學研究所 === 105 === In recent years, there has been an explosive growth in mobile data traffic. Mobile networks are often congested in crowded areas for too many mobile devices connected to the same base stations. In order to address the poor quality of network services in densely...

Full description

Bibliographic Details
Main Authors: Kai-Jie Li, 李凱傑
Other Authors: 蔡志宏
Format: Others
Language:zh-TW
Published: 2017
Online Access:http://ndltd.ncl.edu.tw/handle/h2g9k5
Description
Summary:碩士 === 國立臺灣大學 === 電信工程學研究所 === 105 === In recent years, there has been an explosive growth in mobile data traffic. Mobile networks are often congested in crowded areas for too many mobile devices connected to the same base stations. In order to address the poor quality of network services in densely populated areas, one solution is to deploy a wireless LAN in a high-density way to alleviate network congestion. However, high-density wireless LAN deployments must be properly designed to be effectively improve the quality of network services. Communication between Wi-Fi AP and devices could generate a serious interference to other links without the use of any specific interference control mechanism. There are two main performance indicators concerned in this thesis: the first one is the downlink total transmission capacity in the overall system; the second one is the user fairness index in the overall system, and we propose a complete Wi-Fi AP deployment process, called Capacity-First AP Assignment (CAFAA) method to improve network performance. The process can be divided into two stages: the first stage is the initial AP location and frequency channel assignment phase. We try to reduce the interference in high-density AP deployment through the appropriate choice of AP placement and frequency channel. Then in the second stage we use a well-known optimization algorithm - simulated annealing method, to find a better combinations of AP location and frequency channel by continuous search for new solutions. At last we will show our proposed scheme, CAFAA, could effectively enhance the total transmission capacity in the overall system and meanwhile maintain a certain quality of user fairness.