本論文主要是探討在正交分頻多工存取系統下,限制每位使用者的比例速率及固定總傳輸功率下,利用適應性的資源分配來提高整體容量的演算法之效能改進。正交分頻多工存取系統是由正交分頻多工系統衍生出的多用戶調變技術,每位使用者因為經過不同的傳輸路徑而造成通道響應上的差異,因此利用演算法的技術,可去除通道情況較不好的子載波,而將子載波適當的分配給使用者,再將整體功率做適當的配置以達到每位使用者所需要的速率,並且提升整體傳輸容量。本論文提出一種新的適應性資源配置演算法,首先同時考量各使用者的傳輸速率比及遠近效應等兩種因素分配各使用者子載波數,再以最小傳輸功率為基礎來配置子載波給各使用者。論文中,將此一適應性資源配置演算法與其他三種演算法分別在總傳輸容量及複雜度做比較分析;在總傳輸容量方面,無論是利用向農公式或是以適應性調變的方式來計算容量,此一演算法在各種情況下均較其他三種演算法為優。在複雜度方面,在每位使用者平均分配子載波數多時,其複雜度亦較其他三種演算法低。
This thesis mainly researches into the performance improvement of adaptive resource allocation algorithms in OFDMA systems with both proportional rate constraint and fixed total user’s power. OFDMA system is a multi-user technique derived from OFDM. Since each user has the different channel response in OFDMA systems, we can use algorithms to avoid using the subcarriers with low SNR for each user. Thus, the systems performance can be improved by allocating the subcarriers and distributing power suitably for each user. We proposed a new adaptive resource allocation algorithm in this thesis. First, the number of subcarriers for each user is determined by considering the factors of both proportional rate constraint and near-far effect. Then the subcarriers for each user are allocated based on minimum power criterion. In this thesis, we compare our proposed algorithm with the other three algorithms in capacity and complexity. The total capacity of our proposed algorithm is better than the others either using Shannon theorem or adaptive modulation. When the average number of subcarriers per user is high, our proposed algorithm has lower complexity than the others.