以現在智慧型行動裝置普及率高的情形下,人們可以在任何時間,任何地點進行資料傳輸,使得資料的傳輸量也相當龐大,因此會造成核心網路的負擔。長期演進技術(Long Term Evolution,LTE)提供了一種簡單、省時省力,且低成本的方式進行佈署4G網絡基礎設施,是目前4G標準中被看好的一種網路存取技術,在LTE的環境架構之下,規範了行動用戶(User Equipment,UE)如何透過無線訊號與LTE核心網路(Evolved Packet Core,EPC)的基地台(Evolved NodeB,eNB)連線並進行無線通訊。 為了提供多媒體服務和高頻寬數據傳輸,LTE在MAC層中還具有了Quality of Service (QoS)的概念,並依照資料流量總類的不同而分了不同的QoS等級。如何在有限的無線資源底下,做出最有效的利用且又符合QoS,以及公平性與吞吐量之間的平衡,一直都是排程研究的討論議題。在傳統排程演算法中大多都只是單考慮到rate或delay單一個因素,但其實每一個UE之間所需的rate和delay可能不同,所以本論文將依照各UE之間的rate和delay的不同需求,提出基於rate和delay滿意度之間公平性的演算法,因為考慮到rate和delay滿意度之間公平性,而效能分析的部分,在滿意度的公平性效益上,本文所提出的兩個排程演算比其它對照組演算法還要來得好,與對照組之間的差距最大分別約有50%和44%左右。
Smart mobile devices become more popular in recent years, it is more convenient for people to transmit information at anytime and anywhere, so the amount of data transmission in wireless network would be raised quickly. Long Term Evolution (LTE), which provides a simple and cost-efficient way to deploy the 4G network infrastructure, is one of the radio access technology standards and it is also a candidate standard of 4G mobile data transmission technology. LTE technology is expected to substantially improve end-user throughput to meet rapid growth in demanded mobile application. In LTE network, user equipment (UE) can connect and communicate with the LTE core network (Evolved Packet Core, EPC) according to LTE base station (Evolved NodeB, eNB) with air interface. In order to provide multimedia services and higher data transmission rates, LTE MAC layer has adopted the concept of the Quality of Service (QoS) with different classification service levels. The MAC scheduling is an important research issue, some popular purposes such as how to efficiently allocate the limited radio rescources to have good balance in QoS, throughput and fairness. Many conventional scheduling algorithms only considerd a factor of rate or delay. In fact, all users may have different rate and delay requirements according to different services. In this thesis, the novel scheduling schemes, which can consider the requirement of rate and delay at the same time, are proposed and called MRDUF/FCFS and MRDUF/MT for the LTE downlink environment. Simulation results have shown that the proposed schemes outperform the conventional scheduling algorithms, including MT, PF, BET, and EDF, in term of fairness of satisfaction of rate and delay.