IEEE 802.16e OFDMA TDD 測距程序與整合建構於即時作業系統DSP 平台之上行傳收系統

Abstract

本篇論文首先介紹IEEE 802.16e 測距程序及在我們研究中所使用到的測距演算法。接著，我們將探討如何在數位訊號處理器平台上運用即時作業系統來整合上行傳收系統。 在WiMAX的規格裡，測距是一個很重要的程序。其中，完成功率的調整、時間和頻率偏移的估計、以及基地台和所有用戶之間同步的測距程序，亦稱初始測距。一個行動裝置必須成功地與基地台完成初始測距程序後，才能建立起通訊網路。 在這篇論文裡，我們只探討測距程序中之初始測距的細節及其演算法。然而，其他測距程序，如週期性測距及頻寬要求等議題，將不會在這篇論文裡詳談。 在測距程序裡，基地台會對所有接收到的測距碼作偵測、時間和頻率偏移之估計以及功率程度的量測。然後，基地台以廣播的方式將每組接收到的測距碼以及其所對應的時間、頻率和功率調整資訊回傳。此外，測距程序的成功與否也以廣播的方式告知。 在頻域上，我們大量地遞減測距碼的偵測以及時間和頻率偏移的估計所需的乘法運算量。而與時域上所用互相關演算法比較，我們所提出之演算方式只需要1%的乘法運算量。 在整合上，所實現的上行傳收系統建構於一台個人電腦以及四個Sundance數位處理器模組組成的數位處理器平台。在軟體的開發環境，我們選擇3L Diamond 平台，為3L公司以多處理器而設計及開發的平台系統。 在整合系統上，為了數位處理器的計算效益考量，所有的資料存放格式均是定點數。為了使加速程式執行的速度能夠達到即時處理之需求，我們也對程式做了最佳化處理，如：善加利用數位處理器所提供之函式庫。此外，程式碼的管線化處理及Diamond所支援的特定函式也有助於達到資料的即時處理。 最後，經實驗結果指出，用四個數位處理器模組處理一個上行傳收的架構需要6.5ms。若我們用一個數位處理器模組來完成此工作則須花費17.16 ms。

In this thesis, we firstly introduce the IEEE 802.16e OFDMA TDD ranging process and its algorithms. Secondly, we discuss about how the uplink transceiver system is implemented on digital signal processor (DSP) platform with Real-Time Operating System (RTOS). Ranging is one of significant processes in the mobile WiMAX standard.Power adjustment, timing and frequency offset estimation, and synchronization between a Base Station (BS) and all users within a cell are done through the ranging process, also known as initial ranging. Any Mobile Station (MS) that attempts to establish a communication link is required to carry out a successful initial ranging process with the BS. In this thesis, we discuss about the details of initial ranging and algorithm. In fact, there still exist other types of ranging processes such as: periodic ranging and bandwidth request. However, initial ranging is the only ranging process being discussed in this thesis. In the initial ranging process, BS is required to detect different received ranging codes and estimate the timing, frequency offset and the power level for each user that transmits an initial ranging code. The BS then broadcasts the detected ranging codes with adjustment instructions for the timing, frequency and power level. The status notification of either successful ranging or retransmission is also broadcasted. We accomplish the ranging code detection and the estimation of its timing and frequency offset in the frequency domain. The number of complex multiplications used for ranging code detection and estimation of its timing offset are only 1% of using cross correlation to complete that work in time domain. In the integration work, we implement the uplink transceiver system on the DSP platform which includes a personal computer and four DSP modules (SMT395) made by Sundance. Also, we select Diamond platform as our development environment, which is a 3L’s system for multiprocessor software design and implementation. On the system, only fixed-point data format is supported because of the computational efficiency in DSP. Our optimization goal is to accelerate the speed of program execution time so that it can achieve the requirement of real-time processing. Thus, greatly make use of DSP library function is one of the optimization techniques. Furthermore, software pipelining and several specific functions supported by Diamond also contribute in achieving the real-time processing. Finally, the experimental results show that our system needs 6.5 ms to process an uplink frame, whereas that needs 17.16 ms with only one processor.