Hotspot based Multi-Layer Load Balancing for Improving UE QoS Satisfaction Ratio in Dense Small Cell Networks

碩士 === 國立中山大學 === 資訊工程學系研究所 === 107 === Dense small cell networks were introduced to face the rapid growth in spectral demand of mobile service. However, once the load among the small cell network is unbalanced, the system performance may decrease. One of the concepts of Self-Organized Network (SON)...

Full description

Bibliographic Details
Main Authors: Pei-lun Tsai, 蔡佩倫
Other Authors: Wei-Kuang Lai
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/p2usdr
Description
Summary:碩士 === 國立中山大學 === 資訊工程學系研究所 === 107 === Dense small cell networks were introduced to face the rapid growth in spectral demand of mobile service. However, once the load among the small cell network is unbalanced, the system performance may decrease. One of the concepts of Self-Organized Network (SON) is that the overloaded cell can transfer traffic toward adjacent light-load cell automatically to solve the overloaded problem and make efficient use of spectral resources across the whole network. Most of previous works focus on adjusting the cell range to offload users based on the load difference with the neighbor cells, since most works are decentralized architectures, the authors don’t discuss which overloaded cell can offload its users and which user can be offloaded to a neighbor cell first. However, this problem is worth researching as the remaining resources of the neighbor cells are limited. Therefore, this thesis adopts a centralized architecture, we discuss which user under which overloaded cell can use the idle resources of the adjacent lightly-cell first by collecting the load status of cells and the user resource requirements to maximize the number of satisfied users. In addition, if in hotspot there are many layers of overloaded cells, adjacent cells of overloaded cell still are high-load or even overloaded so we need to start offload traffic from the outermost layer to prepare resources for inner layers. In order to transfer traffic efficiently, we need to let outermost layer cells shift the traffic toward its adjacent cells to release resources first, so that the inner layer cell can have available resources to transfer the traffic toward adjacent cells to solve overloaded problem.