寬頻分碼多重擷取系統之部分平行干擾消除多使用者接收機:演算法及FPGA實現

Abstract

部分平行干擾消除(partial parallel interference cancellation:PIC)是一個簡單且有效率的多使用者偵測技術，此種方法需要部分消除因子(partial cancellation factors :PCFs)作為參數。近年來，可以達到有最小平均平方誤差(minimum mean square error:MMSE)和最小位元錯誤率(minimum bit error rate :MBER)的最佳PCFs值被推導出來，但現有的推導結果都是假設所有使用者的處理增益(processing gain: PG)是相同的。然而在實際的商業系統如W-CDMA系統上並非如此，W-CDMA系統使用可變展頻的方法使每個使用者可以有不一樣的處理增益。在本篇論文中，吾人提出三種最佳的二級partial PIC偵測器來解決這個問題。吾人使用MMSE的方法可以在使用者的PG不同的條件下推導出最佳的PCFs值，事實顯示使用吾人所推導的PCFs值可以增進很大的效能。最後吾人依照FPGA設計流程實現一可應用於W-CDMA系統上低複雜度之接收機，此接收機包括了碼追蹤(code tracking loop)電路。此外吾人並將碼擷取及頻率偏移估測電路和吾人所提出的偵測器整合在一起，使整個partial PIC多使用者接收機更為完整。

The partial parallel interference cancellation (PIC) is known to be a simple yet effective multi-user detection scheme. Parameters need to determine for this approach are partial cancellation factors (PCFs). Recently, optimal PCFs achieving minimum mean square error (MMSE) and minimum bit error rate (MBER) have been derived. Existing results all assume that processing gains for all users are identical. However, this is not true for commercially operated systems such as wideband CDMA (W-CDMA). W-CDMA uses a variable spreading scheme in which all users can have different processing gains. In this thesis, we propose three optimal two-stage partial PIC detectors to solve the problem. Using the MMSE criterion, we are able to derive the optimal PCFs under the different processing gain scenario. It is shown that the detection performance can be greatly enhanced using the derived PCFs. Finally, a low complexity receiver architecture for the W-CDMA system, including the code tracking loop, is proposed and implemented using a FPGA design flow. We also integrate a code acquisition and a frequency offset estimation circuit into the proposed architecture and this results in a complete partial PIC mutli-user receiver.