運用於自我組織網路之多類型Wi-Fi存取點佈建效能最佳化

Abstract

近年來行動網路的興起及智慧型行動裝置的普及，造成有越來越多行動裝置具有連網功能，隨時隨地上網成為習慣；使用者對網路的需求越來越高，但不希望有延遲或無法傳輸資料的情況發生，因此，如何降低干擾或提升使用者的服務品質是一個重要的議題，隨著IEEE 802.11標準中802.11n及802.11ac通訊協定使用2.4GHz與5GHz雙頻基地台的出現，讓此議題有更新的進展。在本篇論文中，我們希望藉由佈建不同種類的基地台(單頻、雙頻或三頻)和配置不同的通訊協定(802.11 g/n/ac)及頻道，達到降低頻道間的干擾和維持使用者的服務品質，也就是讓使用者的延遲能盡量降低。主要希望可以運用於機場基地台之佈建，因為機場是人流最多最密集的區域，各種網路需求都會產生，且不同商家和機場本身就有不同的基地台配置，每個人所使用的手機型號和應用也不同，因此會是干擾最常發生也最需要解決的場合。 我們將上述的配置問題轉化成一數學規劃問題，目標式為最小化使用者的延遲，並提出一個以拉格蘭日鬆弛法為基礎的解題方法，透過鬆弛某些限制式，讓我們可以更快速的獲得可行解，再轉化為原問題的可行解，以求得最佳化的參數運用於配置規劃，其結果可以給機場的管理者和網路服務營運商作為佈建基地台的參考。

With the traffic demands are increased in 802.11 wireless networks, more and more mobile devices have applications connected to the Internet. Under this trend, the delay situations and unstable transmission environments are not satisfied on data rate requirements. Therefore, how to reduce the interference and improve the user's Quality of Service (QoS) is an important issue. With the appearance of protocol using 5GHz band and appearance of dual-band and tri-band APs, let this issue having new development. In this thesis, we consider through the placement of the APs and choose different types of AP, each of which have to select corresponding protocol, and channel assignment to reduce co-channel interference and maintain user’s QoS, namely enable user’s delay can lower as far as possible. We mainly hope the AP assignment algorithm can be used in airport. Because airport is the most populous places, many demand of internet will happened. Stores in airport and airport itself will have different AP configuration, this may cause more interference. Therefore, airport will be most interference place and must be solved. We apply mathematical model to describe the above assignment problem. Objective function will be minimum user’s delay. A Lagrangean Relaxation (LR) based method will be proposed to solve the problem. Through relaxing some constraints, enable us to obtain solution more quickly. Than transform the solution to solution of primal problem, getting the optimal parameters for the configuration planning. The result can be a reference of AP assignment for manger of airport and Internet Service Providers (ISP).