多層次資料庫系統應用在個人通信服務的效能評估研究

Abstract

在本論文中，我們提出了應用在多層次個人通信服務系統 (multi-tier PCS systems) 中可能的兩種多層次資料庫系統(multi-tier database systems)架構，分別是集中式 的架構(Centralized HLR Architecture, CHA)和非集中式的架構(Decentralized HLR Architecture, DHA)。 我 們也提出了在這兩種架構中如何管理以及處理和移動有關的 運作，亦 即註冊(registration)、取消註冊(deregistration)、呼叫的撥接(call origination)以及呼叫的遞送(call termination)。在一些假設下，我們 導出了對於 系統效能評估的數學分析式，評估的項目包括查詢所需的 反應時間(query response time)、更新所需的反應時間(update response time)以及在這多層次資料庫系統網路 中傳輸的信號量( traffic volume)。最後我們經由分析結果來比較這兩種架構的優劣。 從 我們的分析結果可得知，除了在更新的反應時間這一項以外，非集中式架 構(DHA)的 效能要比集中式架構(CHA)來的好。 In this thesis, we propose two possible multi-tier database system architectures, Centralized HLR Architecture (CHA) and Decentralized HLR Architecture (DHA) for multi-tier PCS systems. We also propose mobility management strategies which deal with mobility-related operations (i.e., registration, deregistration, call origination, and call termination) for the two proposed architectures. Under some assumptions, mathematical analysis are derived on the performance metrics, which include the query response time, the update response time, and the traffic volume produced in the two architectures. Finally, the analysis results of CHA and DHA are made, and the pros and cons of the two proposed architectures are figured out. From the analysis results, it is found that the DHA performs better than the CHA except the update response time.