由於高解析度媒體的快速發展刺激了對無線寬頻網路的需求以及對於網路連線品質的要求。然而，第四代的無線寬頻網路WiMax (Worldwide Interoperability for Microwave Access) 和 LTE/LTE-A (Long-Term Evolution /Long-Term Evolution-Advanced) 的技術提供了更高速的數據傳輸以及更大容量的無線接取網路，由於各種行動裝置應用使得用戶通過手持設備訪問相關聯的服務，因此節能已成為行動設備中最關鍵的問題之一，尤其是在實際應用中增加其操作時間。 雖然WiMax和LTE / LTE-A網絡提供各種睡眠模式類型和參數以實現節能，但是當BS (Base Station)或eNodeB (evolved Node B)分配無線電資源時，可以通過考慮更多的影響因素來改進睡眠模式的效率，而且不管在最佳化能源的利用率或是滿足QoS (Quality of Service)的需求，在各種無線寬頻網路仍然需要新穎的無線電資源分配管理方法。為了達到上述的目標，本論文提出ESC-US (Energy-Saving Centric Uplink Scheduling) 的方法在WiMax上行鏈路中使用以及 SDRRS (State Dependent Radio Resource Scheduling)的方法在LTE下行鏈路中使用。在ESC-US的方法中提出了rtPS (real-time Polling Services)和nrtPS (non-real-time Polling Services)兩種分配頻寬的演算法，此研究提出了讓用戶先進入睡眠模式在能滿足QoS的要求下再醒來傳輸資料，用來達到節能的目的。在模擬的結果中此方法比起傳統的方式能夠滿足用戶的QoS要求下並且達到更好的節能效果。在SDDRS的方法中加入了burst-scheduling 的觀念去傳輸資料，並且在醒來和休眠的狀態中去調整inactivity timer的值，從模擬的結果來看節能的效能也是優於其他的方法。 在LTE-A網絡中，具有LTE-A能力的UE (User Equipment)可以聚合多個CC (Component Carriers)以用於使用更多的頻寬。但是具有LTE-A能力的UE可以被分配到在不同CC上的RBs (Resource Block)，LTE能力的UE只可以在同一個CC上被分配到RB。為了解決此問題，本論文提出了一個使用Gale-Sharply 的方法用來分配CC的方式，讓不論是LTE的UE或是LTE-A的UE可以在LTE-A的網路環境分配上行鏈路的排程方法。 關於機器與機器之間通訊的其他研究，其研究上行鏈路調度方案以最小化信令開銷並且最大化LTE網絡中的系統吞吐量為目的，在機器與機器通訊中的關鍵問題是減少大量連接的設備和eNodeB之間的信令開銷，本論文提出方案動態地調整群組成員，不僅考慮QoS和通道條件，而且應用分配請求概念來分配剩餘頻寬。 ;The rapid development of high-definition media has stimulated demand for wireless broadband networks and connection-oriented quality for several network services. However, the fourth-generation standard, WiMax (Worldwide Interoperability for Microwave Access) and LTE/LTE-A (Long-Term Evolution /Long-Term Evolution-Advanced) technologies, provide a higher-speed data rate and larger-capacity wireless access networks. Because various mobile applications induce users to access the associated services through handheld devices, energy saving has become one of the most critical concerns in mobile devices, especially for increasing their operation time in practical applications. WiMax and LTE/LTE-A networks provide various sleep mode types and parameters to achieve power-saving, it can be improved by considering more influence factors when BS (Base Station) or eNodeB (evolved Node B) allocated radio resource. However, new radio resource scheduling is still needed to consider either optimal energy utilization or QoS (Quality of Service) satisfaction for both broadband access networks. To achieve this objective, we propose an ESC-US (Energy-Saving Centric Uplink Scheduling) scheme for uplink traffic in WiMax and SDRRS (State Dependent Radio Resource Scheduling) scheme for downlink traffic in LTE. The ESC-US scheme provides rtPS (real-time Polling Services) and nrtPS (non-real-time Polling Services) scheduling algorithms that apply the “just enough QoS” and “sleep before transmission” concepts to achieve energy-saving centric objective. The simulation results demonstrate that both schemes satisfy the desired QoS and achieve better energy consumption compared to the conventional scheme. The SDRRS scheme involves the burst-scheduling concept with respect to transitions between the active and the sleep states to adjust the inactivity timer. The performance was investigated via exhaustive simulations indicated that the SDRRS scheme can reduce energy consumption more effectively compared with the other schemes. In LTE-A network, the UE (User Equipment) with LTE-A capability could aggregate more than one CC (Component Carriers) for more channel bandwidth. However, the LTE-A UE can be allocated RBs (Resource Block) in different CCs, but LTE UE can only be allocated RBs in the same CC. To address this issue, we propose a novel scheme that considers the localized subcarrier by properly using the Gale-Sharply concept to schedule the uplink traffic of both LTE and LTE-A UE in LTE-A network environment. Additional research is presented on machine-to-machine communication, which studies an uplink scheduling scheme to minimize signaling overhead and maximize system throughput in LTE network. A crucial problem in the machine-to-machine communication is to reduce signaling overheads between a large number of connected devices and eNodeB. We propose scheme dynamically adjusts group members that considers not only QoS and channel condition but also applies the allocation-before-request concept to allocate the residual bandwidth.
