本論文針對異質網路之系統下，考慮使系統傳輸量最大化的問題。在第四代行動通訊中，異質網路被視為是一個主要的方案；由於所有基地台使用相同的頻段，細胞間的干擾會是個重要的問題。因此在增強小區間干擾協調技術(eICIC)中，近乎空白子時框(ABS)的技術被提出解決干擾的問題。增強小區間干擾協調技術主要的特性是某些子時框只允許被特定的使用者使用，因此在增強小區間干擾協調技術下去做資源配置是很困難的而且與傳統的資源配置問題不同。在本篇論文中，資源配置的問題包含了架構近乎空白子時框的模型以及使用者間的資源配置。 在本篇論文將會分成兩個部份探討：第一個部分，在時域部分利用通道的特性去設計近乎空白子時框的模型；論文的第二部分，利用Lagrangain技術發展一個使用者分群的策略，利用這個策略決定使用者要被分配到正常的子時框還是被保護的子時框。此外在本篇論文中，針對頻域部分考量使用者間傳輸量公平性，因此資源分配的問題都是以使用者間的公平性為基礎。 本論文提出的方法同時考慮系統的效能以及使用者間的公平性，而且相較於最佳解方法有著較低的計算複雜度。在模擬結果顯示出本篇論文的方法勝過於現存的演算法。 In this thesis, the problem for maximizing the system sum rate in heterogeneous networks (HetNet) is considered. HetNet layout including macro eNodeBs and some low-power nodes are considered as a key scenario in LTE-Advanced. The major problem in HetNet is the co-channel inter-cell interference because the spectrum is shared between all eNBs. In order to resolve the interference problem, the utilization of almost blank subframes (ABS) in time domain enhanced inter-cell interference coordination (eICIC) is introduced. The main feature of time domain eICIC is that some subframes are only allowed to be used by some particular user equipment (UEs). Therefore, eICIC makes the resource allocation problem difficult and different from conventional resource allocation problems. The main challenges of the resource allocation problem in this thesis include the configuration of ABS pattern and the resource allocation for macro UEs and pico UEs. The first part of this thesis introduces an ABS pattern design by using the channel characteristic, which is developed in the view of time domain. After ABS pattern is configured, the second part of this thesis develops a UE grouping strategy based on the standard Lagrangian technique to determine which pico UEs use the protected or normal subframes having co-channel inter-cell interference. Besides, the issue of proportional fairness among UEs is also studied. The resource allocation problem in the view of frequency domain is considered, especially for pico UEs. The assignment of resource blocks is developed based on the fairness among UEs. The proposed schemes take the system performance and the fairness into account, and has lower complexity compared to the full search algorithm. Simulation results also reveal that the proposed schemes outperform the existing algorithms.
