移動管理在行動網路中是重要的功能，其主要任務為掌握使用者的所在位置，當有通話時協助通知 (Paging) 用戶相關事件。為了明確掌握用戶的位置，每當用戶移動到新的基站範圍時，須對系統發送位置更新 (Tracking Area Update)的請求。高速移動的用戶勢必產生頻率較高的位置更新請求。目前行動通訊系統以群組基站範圍為大區域的方式降低用戶發送位置更新的成本，其缺點為通知時需讓整個區域內所有基站一起進行通知 (Paging)。因此，如何從位置更新的成本與通知用戶的成本中找到適合的網路規畫一直是有趣的研究課題。過去，移動管理成本的研究已有許多論文提出相關的研究成果，透過不同的機制優化移動管理的功能，例如依照使用者的特性動態地配置追蹤區域 (Tracking Area)。大多數研究圍繞於使用者移動特徵，而較少關注以使用者的通話行為模式。換言之，當使用者的通話頻率增加時才有可能對系統效能造成負面的影響，如何在使用者移動性與通話行為模式中找到平衡點正是本篇論文的目標。 本論文提出了一個位置管理機制的架構，將使用者的通話流量視為考量的因素，並依照使用者的特性配置專屬於該使用者的追蹤區域，利用參數估計以評估使用者的通話行為模式，配合使用模糊邏輯控制器將使用者的機動性以及通話行為模式作為輸入，經由評估後決定出適合於該使用者的追蹤區域。 ;Mobility management is one of the major functions of cellular network. The aim of mobility management is to track the location of UE (User Equipment). When new call arrives, the cellular network has to determine the base station to page the UE according to its location. To do it, UE has to issue tracking area update (TAU) request to cellular network whenever it moves into range of new base station. It is a tradeoff between the paging cost and tracking area update cost. Regarding location management in PCS (Personal Communications Service) network, many location management methods have been proposed and they could be classified into several categories, distance based, time based, and movement based, all these methods consider only the mobility of UEs. The allocation of TA (Tracking Area) is based on the mobility characteristic of UEs. However, we found that the telephonic traffic is another potential factor, which may directly impact the performance rather than the mobility. This thesis proposes an efficient scheme, which treats user′s telephonic traffic as an essential factor, to allocate proper TA for specific UE. We first estimate the parameter of telephonic traffic and then use fuzzy logic controller to decide size of tracking area. Based on it, we can find a TA allocation strategy to minimize the signaling cost of paging and TAU.
