| 摘要: | 近幾年來，寬頻無線網路發展是為了讓使用者能夠無所不在的使用寬頻網路的服務，為了達到能夠隨時隨地使用寬頻網路，行動裝置的電量消耗也是一個重要的問題。 在LTE環境下，下行鏈路方向(downlink) ，eNoed B 分配頻寬的時候屬於即時性的了解 UE 目前的需求資訊，傳統在排程機制設計上，eNode B會根據UE當下的狀態以及目前所擁有的頻寬，分配給每個處於醒著(active or light sleep)的UE，並沒有將省電機制考量進來，導致eNode B在分配資源給UE的時候，往往會超過服務品質要求，造成電量不必要的浪費。因此，本論文在LTE環境下處於DRX模式下提出一個在下行鏈路方向的排程方法，不僅考慮到電力消耗的問題，同時兼顧到不同服務類型的服務品質要求，如延遲容忍(Delay tolerance)等；此外，在頻寬分配上也一併將通道狀況(Channel condition)納入考慮，以提高無線電資源的使用效率。 本論文模擬在 LTE網路的環境下，具有 HTTP 和 Video Streaming 這兩種不同的服務類型時，比較不同排程方法的效能。本論文所提出的方法，與只考慮通道狀況做排程的方法及只考慮delay的方法做比較，我們的方法對於 HTTP而言，在皆能滿足QoS delay budget的情況下，雖然省電的效果差距不大，但能使在不同通道狀況下的 UE 耗電量更具有公平性；對於 Video Streaming 而言，卻能達到比較好的省電的效果並降低delay達到公平性，而且能夠在相同的頻寬，接受更多的用戶。 In recent years, mobile broadband wireless access networks have been rapidly evolved for the increasing demands of users. Because of the battery limits of ,mobile devices, power saving is an important issue for users to access vast range of services anytime, anywhere. For most previous downlink scheduling research, eNode B only uses the UEs’ current packet state (e.g. delay), current channel quality to determine how to allocate resources. However, because power saving for UEs is not considered, these scheduling mechanisms will lead some UEs to exceed their quality of service requirements, and result in wasting power. In this thesis, scheduling mechanism simultaneously considering channel state condition, QoS, and also power saving effect in LTE with DRX environment is proposed.. The basic idea behind the our proposed scheduler is to reduce the useless duration for power consumption (e.g. rx with no data ) as much as possible in order to save power. In this study, we take two different types of services, HTTP and Video Streaming, to simulate the different scheduling methods in the LTE network, and compare the performance of them. The results show that while considering different channel conditions, the proposed mechanism with HTTP request can achieve more fairness than traditional methods. Moreover, for Video Streaming traffic, the proposed mechanism can not only achieve power saving but also serve more users in single eNode B, ( ie. higher throughput )while meeting the UE’s QoS requirement. |
