查詢結果分析
相關文獻
- M準位正交振幅調變型態識別之研究
- 論混合常態之可識別性
- Performance Analysis of QAM Signal Recognition Based on Theoretical Approach over Nakagami Fading Channel
- 多重專家應用於線上手寫中文文字識別
- The Survey of On-Line Chinese Character Recognition
- Seal Identification Using the Delaunay Tessellation
- 從府城到世界的建築與城市觀察-- (40) 文化資產識別系統篇
- 彩色切片影像之自動影像分析系統:鴨肝切片
- 隨機遞減法在建築物微振時間域系統參數識別之應用
- 泛歐式數位行動電話之通道編碼技術
頁籤選單縮合
題 名 | M準位正交振幅調變型態識別之研究=Study of Classification of M-Ary QAM Signals |
---|---|
作 者 | 牛德真; 林家傑; | 書刊名 | 中華工商專校學報 |
卷 期 | 15 1997.09[民86.09] |
頁 次 | 頁79-95 |
分類號 | 312 |
關鍵詞 | 正交振幅調變; 識別; 機率密度函數; QAM; Classification; Probability density function; |
語 文 | 中文(Chinese) |
中文摘要 | 本文提出M-ary QAM信號之分類器,以信號特徵為基礎,分兩階段識別。第一階段 ,取每筆信號資料在四個相位區間內所有信號點的振幅動差作為特徵。以〞最小錯誤率準則 〞作為識別的基礎, 將 M-ary QAM 信號分成〞 M 為 2 的奇數次方〞與〞 M 為 2 的偶數 次方〞兩大類。 第二階段,以 M-ary QAM 信號振幅之機率密度函數為特徵,推導出分類臨 界值,再依此辨認 16QAM、64QAM 或 32QAM、128QAM。 模擬結果顯示,在第一階段分類中,當 SNR 大於 10dB 時,錯誤率在 0.3 %以下,在第二 階段分類中,當 SNR 大於 10dB 時,錯誤率在 1.35 %以下。 |
英文摘要 | This paper pertains to the classification of M-ary Quadrature Amplitude Modulation signals (M-ary QAM). The modulation schemes concerned are: 16QAM, 32QAM, 64QAM, and 128QAM. The classification procedure is composed of two phases. At the first phase, the four QAM signals different in Mare split into two classes: 16QAM and 64QAM(M=2 ��;n is even ) as one and 32QAM and 128QAM (M=2 ��:n is odd ) as the other, the amplitudes whose phases are located in the four intervals: (45 ± R ),(135 ± R),(225 ± R) and (315 ± R ) in degrees are selected and then the sampling statistical moments of the amplitudes are chosen as the features. The purpose at this phase is to recognize the received signals in one of two classes by using the minimum error probability criterion. At the second stage, each class is further split into two subclasses. The probability density functions of amplitudes of signals of signals are employed to derive the required thresholds, which is in turn utilized to classify the signals in the subclass. Simulation results show that the proposed two-stage. Classification algorithm works well when SNR(signal-to-noise) is larger than 10 dB and fair below 10 dB. |
本系統中英文摘要資訊取自各篇刊載內容。