類神經網路應用於多個節點之IEEE802.11無線網路的速率調整
Abstract
我們研究在IEEE 802.11 無線區域網路(WLAN)中有多個工作站使用不同傳輸速率時的吞吐量(throughput)效能。無線網路在實體層(PHY)使用不同的調變與編碼技術(Modulation and Coding Schemes， MCS)支援多重傳輸速率。當某一個工作站的通道品質變差時，它會使用較低的傳輸速率來獲得較佳的連線品質。然而，IEEE802.11 媒體存取控制層(Medium Access Control， MAC)所使用的分散協調式功能(Distributed Coordination Function， DCF)對於每一個工作站提供了相同的傳輸機會，無論它們所使用的傳輸速率為何。這將導致使用高速率的工作站其資料傳輸率被限制在某一個使用中的最低速率，使得整體系統的吞吐量(Throughput)下降。這種現象稱為IEEE 802.11無線區域網路的傳輸效能不規則(performance anomaly)，而這個現象在提供高範圍傳輸速率的網路，例如IEEE 802.11g 網路中將會是更為嚴重的問題。
本研究提出一種調整Auto Rate Fallback（ARF）的演算法，以提高總吞吐量的效能在IEEE 802.11無線區域網路（WLAN）的多個節點。當節點的數量增加，並且封包傳送過程中碰撞的情況也因而增加，使用ARF演算法將有可能降低傳輸速率，所以造成整體的吞吐量相對減少許多。我們提出一種以類神經網路為基底來調整ARF的演算法達到提高吞吐量性能透過動態地調整系統參數，我們根據傳輸速率的競爭情況包括網路節點的數量及網路流量的飽和度來決定系統參數。
We investigate the throughput performance of IEEE 802.11 Wireless Local Area Networks (WLAN) with multiple stations using different data rates. IEEE 802.11 WLAN PHYsical Layer (PHY) provides multiple data rates with different Modulation and Coding Schemes (MCS). If one station experiences bad channel conditions, it would degrade the data rate to achieve a more reliable transmission quality. However, 802.11 Distributed Coordination Function (DCF) protocol at Medium Access Control (MAC) layer essentially provides equal transmission opportunities to each transmitting station. Thus, the throughput of stations with high data rates will be restricted within the lowest rate used by some stations, resulting in the degradation of system throughput. Such the phenomenon is so called “performance anomaly” and can be an extremely severe problem in the network which provides large-scale PHY rates, e.g. 802.11g WLAN with rates ranging from 1 Mbps up to 54 Mbps. The study presents an adaptive Auto Rate Fallback (ARF) scheme to improve the performance of aggregate throughput in IEEE 802.11 Wireless Local Area Network (WLAN) with multiple nodes. When the number of contending nodes increases, using ARF will be likely to degrade transmission rates due to increasing packet collisions and can consequently cause a decline of the overall throughput. In this paper we propose a neural-network based adaptive ARF scheme which improves the throughput performance by dynamically adjusting the system parameters that determine the transmission rates.
We investigate the throughput performance of IEEE 802.11 Wireless Local Area Networks (WLAN) with multiple stations using different data rates. IEEE 802.11 WLAN PHYsical Layer (PHY) provides multiple data rates with different Modulation and Coding Schemes (MCS). If one station experiences bad channel conditions, it would degrade the data rate to achieve a more reliable transmission quality. However, 802.11 Distributed Coordination Function (DCF) protocol at Medium Access Control (MAC) layer essentially provides equal transmission opportunities to each transmitting station. Thus, the throughput of stations with high data rates will be restricted within the lowest rate used by some stations, resulting in the degradation of system throughput. Such the phenomenon is so called “performance anomaly” and can be an extremely severe problem in the network which provides large-scale PHY rates, e.g. 802.11g WLAN with rates ranging from 1 Mbps up to 54 Mbps. The study presents an adaptive Auto Rate Fallback (ARF) scheme to improve the performance of aggregate throughput in IEEE 802.11 Wireless Local Area Network (WLAN) with multiple nodes. When the number of contending nodes increases, using ARF will be likely to degrade transmission rates due to increasing packet collisions and can consequently cause a decline of the overall throughput. In this paper we propose a neural-network based adaptive ARF scheme which improves the throughput performance by dynamically adjusting the system parameters that determine the transmission rates.
Description
Keywords
吞吐量, 無線區域網路, 802.11 WLAN, throughput performance