- 學位論文
非正交多工存取通訊系統之接收機設系與實現
Receiver Design and Implementation of Non-orthogonal Multiple Access (NOMA) Communication System
摘要
非正交多工存取(Non-orthogonal multiple access, NOMA)基於OFDM的載波上，再分配功率並疊加兩個以上用戶的訊號，能提升系統總傳輸率，有望成為下一世代行動通訊採用的多工存取技術。在本論文中，著重在NOMA接收端的設計與實現，並使用維也納大學提供的Vienna LTE simulator作為開發平台。 本論文第四章透過比例公平排程模擬，包括單輸入單輸出和多用戶波束成型系統，證實NOMA相比OMA都有更高的總傳輸率。在NOMA系統下，近用戶端需要先扣除功率較大的遠用戶訊號，再解自己的訊號，我們比較碼字連續干擾消除和符元連續干擾消除兩種強干擾扣除方法，並模擬文獻[30]提出的聯合調變和解調變方法，其中碼字連續干擾消除的效能能接近理想干擾扣除。 本論文第六章結合NOMA和SUMIMO系統，比較各種偵測器的效能和複雜度。在多層訊號的系統下，忽略一點干擾都會嚴重影響非線性偵測的效能，因此我們偵測遠用戶訊號會同時考慮近用戶的弱訊號以提升效能。我們提出限制搜尋範圍方法以降低球面解碼的複雜度，模擬結果顯示可以在維持相同效能下，減少15-70%的複雜度。我們再提出自動切換偵測器，會根據編碼方案、功率因子、通道狀況，切換合適的偵測器，達到錯誤率夠低又複雜度較低的偵測結果。 本論文第七章延伸第六章的內容，結合NOMA和MUMIMO with multi-stream系統。在接收端，我們需要處理波束間干擾。也同樣加上限制搜尋範圍法和自動切換偵測器以降低複雜度。
並列摘要
Non-orthogonal multiple access (NOMA) which superposes multiple users’ signals in the power domain based on OFDM waveform, can improve system throughput. NOMA is one of the promising radio access techniques for next mobile communication generation. In this thesis, we focus on receiver design of NOMA, and we use Vienna LTE simulator as simulation environment. In chapter 4, by proportional fair scheduling, we verify that NOMA has larger total throughput than OMA, both in SISO and multi-user beamforming system. In NOMA system, near user needs to cancel stronger far user’s signal first, and then detects its signal. We compare codeword successive interference cancellation (SIC) and symbol SIC, and implement jointly modulation and detection [30]. We show that codeword SIC can approach the performance of ideal SIC. In chapter 6, we combine NOMA with SUMIMO; compare performance and complexity of various detectors. For multi-stream signal, even weak interference would seriously affect non-linear detector, so we consider weaker near user’s signal to detect far user’s signal. We propose limited search space method to reduce complexity of sphere decoder. By simulation, we show that it can reduce 15-70% complexity and maintain the same performance. We also propose adaptive detector, which can switch detector according to MCS, power factor, and SNR, so as to achieve good enough error rate and low complexity. In chapter 7, we combine NOMA with MUMIMO with multi-stream, and we need to deal with IBI at receiver side. Also, we implement limited search space method and adaptive detector to reduce complexity.