| 摘要: | 超寬頻(Ultra-wideband)主要用於室內環境，和多天線系統的結合可以提供更高的傳輸效能，但環境會產生多路徑效應(multi-path effect)。由於此效應造成的符際間干擾(Inter Symbol Inference, ISI), 使得通訊位元錯誤率及失效率(outage probability)增加, 導致通話品質變差。因此在本論文中，透過使用實數正交設計(Real Orthogonal design, ROD)可以有效的改善超寬頻多天線系統，將模擬環境透過射線彈跳追蹤法結合反複利葉轉換可以求出該環境的時域脈衝響應，並使用耙式接收機(RAKE receiver)增加接收信號強度以抑制多路徑效應。在本篇論文中，我們將針對天線的排列方式(CASE A)和發射端的高度與演算法搜尋之位置(CASE B)進行探討無線存取點在環境中之最佳位置。CASE A以線形、L形、圓形、Y形四種不同類型的8 x 8發射和接收天線陣列，探討在3.1~10.6GHz之頻段時的超頻寬通訊(Ultra-wideband, UWB)之性能。數值結果顯示，在一般環境中，圓形的排列方式的失效率(定義為錯誤率>10-6)，在大部分的情況下，是優於其他三種(線形、L形、Y形)排列方式的。CASE B 根據已知的室內環境以及所選定的UWB-MIMO系統，藉由隨機式全域最佳化演算法找到發射端在環境中之最佳位置，並分別調整發射端的高度，其中以非同步粒子群聚最佳化法（Asynchronous Particle Swarm Optimization, APSO），所找到之最佳位置，效果比自我適應之動態差異型演化法(Self-Adaptive Dynamic Differential Evolution, SADDE)來的更好、更迅速。本研究在學理與實際之通信系統規劃上，皆具有其重要性。
Ultra-wideband (UWB) combined with multiple input and multiple output (MIMO) are proposed to increase the data rate. UWB mainly used for indoor environments. Due to the multi-path effect, the Inter Symbol Inference (ISI) increases the bit error rate and outage probability of the MIMO-UWB system. In this thesis, the application of Real Orthogonal design is employed to improve MIMO-UWB system. The ray-tracing technique and inverse fast Fourier transform are used to get the impulse response of the indoor environment. Moreover, RAKE receiver is used to increase the strength of received signal against multiple path effect.
In this thesis, different antennas array (case A) and the best position for different height of transmitter are presented (case B). There are four different shapes of arrays for the transmitter and receiver, such as linear array, L array, circular array and Y array, for CASE A. The performance for these array in UWB frequency between 3.1~10.6GHz are discussed. The outrage probability of circular array is better than the other three arrays. For Case B, two antenna element arrays are used for the transmitter with different height. The optimal location for the transmitter is searched by Asynchronous Particle Swarm Optimization (APSO) and Self-Adaptive Dynamic Differential Evolution ( SADDE). The numerical results show that the performance for APSO is better than SADDE. |
