Full metadata record
DC FieldValueLanguage
dc.contributor.author王啟宇en_US
dc.contributor.authorWang, Chii-Yeuen_US
dc.contributor.author楊啟瑞en_US
dc.contributor.authorYuang, Maria C.en_US
dc.date.accessioned2014-12-12T02:40:32Z-
dc.date.available2014-12-12T02:40:32Z-
dc.date.issued2013en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#GT070056501en_US
dc.identifier.urihttp://hdl.handle.net/11536/74429-
dc.description.abstract在IEEE 802.11的規範中,支援多種不同的傳輸頻寬。在基礎架構模式(Infrastructure mode)下,當所有裝置處於同一個基本服務集(Basic Service Set, BSS),因為距離無線網路基地台(Access Point)較遠等原因,裝置的傳輸頻道發生干擾較嚴重的情形時,會自動選擇頻寬較低的調變類型,降低傳輸速度,提高封包傳送成功率。在目前已知的研究中已提出,當多個裝置以高速的頻寬傳輸時,只要存在一個裝置以低速頻寬傳輸,就會降低其他裝置的傳輸效率,導致其他高速率傳輸裝置的吞吐量(Throughput),會降低至與低速率傳輸裝置吞吐量相同的數量級。而在IEEE 802.11的研究中,通常稱此現象為「性能異常」(Performance Anomaly)。 IEEE 802.11n自2009年發表以來,相關支援的Wi-Fi認證產品已非常普及。而本篇論文主要探討,在IEEE 802.11n的規格參數設定下,在飽和(Saturation)與非飽和狀態下,以調整最小競爭視窗大小(minimum contention window size,簡稱CW_min),改善異常現象之效果。 在研究的過程中,我們先證明以自行撰寫之模擬程式,可以模擬出性能異常現象。再邏輯性的去實驗解決此問題之方法,以分別調整頻寬、流量、最小競爭窗框大小、共同競爭裝置數等參數之後,討論對於高傳輸速率之裝置有何影響。實驗結果指出,當裝置數夠多的時候,將高傳輸速率裝置的CW_min設定為最小值,低傳輸速率裝置的CW_min設定為最大值,可以讓高傳輸裝置的效能,在吞吐量、延遲時間等方面的數據結果達到最明顯的改進。zh_TW
dc.description.abstractThere are many different specifications of transmission rate which are set by IEEE 802.11 standards. Under the infrastructure mode, mobile stations which are far away from the AP, will suffer much more signal interference. In this circumance, those stations will change their transmission rate to a lower value. It has been known that, the existence of the low rate stations will affect the other high rate stations severely. In the related research work of IEEE 802.11, we often refer it as performance anomaly. Since the IEEE 802.11n standard had been released in 2009, it has been widely applied among many products around the world. The purpose of our work is to solve the performance anomaly problem by tuning the minimum contetion window size(CW_min), under the 802.11n specifications. First, we prove that the self-developed simulation program can simulate the abnormal phenomenon mentioned above. Second, we make experiments with different parameters such as transmission rate, traffic arrival rate, CW_min and number of contending stations. And observe the variations of throughput, mean delay and delay jitter. Throughout our experiments, we show that when the number of contending stations are large enough, maximizing the CW_min of the low rate stations and minimizing the CW_min of the high rate stations highly improved the performance of the high rate stations.en_US
dc.language.isozh_TWen_US
dc.subjectIEEE 802.11zh_TW
dc.subject基礎架構模式zh_TW
dc.subject無線網路zh_TW
dc.subject競爭視窗zh_TW
dc.subject吞吐量zh_TW
dc.subjectIEEE 802.11en_US
dc.subjectMinimal Contention Window Sizeen_US
dc.subjectTransmission Rateen_US
dc.subjectThroughputen_US
dc.subjectDelayen_US
dc.titleIEEE 802.11n性能異常現象之改進zh_TW
dc.titleImprovement of IEEE 802.11n Performance Anomalyen_US
dc.typeThesisen_US
dc.contributor.department網路工程研究所zh_TW
Appears in Collections:Thesis