無線傳輸之跨層最佳化佇列分析

Abstract

近年來，自適應調變及編碼技術被廣泛應用在無線通訊系統上。若不考慮系統本身是否有充足的緩衝空間及足夠大的超時限制，自適應調變及編碼技術能被應用在系統上滿足最大吞吐量。然而，這樣的假設是不實際的，當系統的緩衝空間及超時限制為有限時，在網路上的佇列效應可能會對系統的吞吐量造成極大的影響。在這篇論文中，我們分析了考慮跨層及自適應調變及編碼技術的機制，並且將無線通訊系統塑造成一個泊松程序及服務速度由馬可夫調變的佇列。我們考慮擁塞及超時造成的影響，重新定義了系統的吞吐量，並且得到最佳的設定去取得系統的最大吞吐量。我們設計了幾組不同的系統參數所造成的結果。

In recent years, adaptive modulation and coding (AMC) schemes are widely implemented in wireless communication system. Assuming the system always has sufficient buffer size and the time-out constraint are not taken into account, AMC can be employed to achieve the maximal throughput at physical layer. However, this assumption is invalid in practice. The queue effect at higher layer may be influential in throughput when the buffer size is limited or considering the time-out constraint. In this thesis, we analyze the cross-layer AMC scheme by modeling the wireless communication system as a Markovmodulated Poisson process (MMPP)/D/1/Markov-modulated service speed (MMSS) queue. We redefine the throughput with considering the blocking and time-out issues, and obtain the optimal setting to achieve the maximal throughput. Numerical results illustrate the dependence of system performance on different system parameter settings.