個人行動通訊網路中預測式移動管理策略

Abstract

在個人行動通訊網路中，網路系統利用移動管理讓手機能在網路涵蓋的區 域內自由的移動。因為個人行動通訊的用戶不斷增加，手機移動管理的成 本也隨之增加。在本篇論文中，我們將針對個人行動通訊系統的移動管理 ，提出一個前瞻位址的預測策略，來降低移動管理的成本。此一預測式策 略是根據手機過去移動與被叫受話的行為模式所發展出來的。我們利用電 腦模擬所獲得的結果顯示，此一預測式策略優於現有策略。雖然此一策略 需增加中央處理器的運算，但中央處理器運算能力和資料庫容納能力隨科 技不斷增強，而無限頻寬資源卻仍是有限的，所以用此一策略來節省無限 頻寬資源是可以接受的。 Personal communications services (PCS) network uses mobility managementprocedures for supporting mobile terminals free to move within networkcoverage area. Because the number of PCS mobile subscribers keeps increasing, the network impact for keeping track of location of mobile terminals wouldbe acute. In this thesis, we focus on reducing costs of mobility managementin PCS networks. A predictive strategy which aggressively registers thelocation of mobile terminals to "look ahead" location according to mobileterminal's mobility and call arrival pattern is proposed. Simulation resultsshow the effectiveness of the proposed predictive strategy. Although ourstrategy needs more CPU processing cost. But it is acceptable that usesincreasing CPU computational power and database capacity in exchange forlimited wireless resources.