A Hht-Based Time-Frequency Features Analysis for Underwater Acoustic G.729 Voice Signal-3

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

NSC101-2221-E022-005

Translated Name/計畫中文名

一種水中聲學 G.729 語音通訊訊號希爾伯特黃時頻特徵解析技術-3

Project Coordinator/計畫主持人

Chin-Feng Lin

Funding Organization/主管機關

National Science and Technology Council

Co-Investigator(s)/共同執行人

張順雄
文展權

Department/Unit

National Kaohsiung University of Science and Technology, NKUST

Website

https://www.grb.gov.tw/search/planDetail?id=2585921

Year

2012

Start date/計畫起

01-08-2012

Expected Completion/計畫迄

01-07-2013

Bugetid/研究經費

674千元

ResearchField/研究領域

電子電機工程

在這個計畫，我們將探討 G.729 水中聲學語音通訊訊號希爾伯特黃時頻特徵。我們將壓縮之 G.729 語音位元串流，輸入我們所發展之 Matlab 水中聲學多輸入多輸出 OFDM 多媒體通訊軟體平台(NSC-99-2221-E-022-019)，輸出所接收之 G.729 水中聲學語音訊號位元串流，並解壓縮。所接收和解壓縮之語音訊號，應用經驗拆解法，拆解成數個本質模態函數和一個殘餘函數。並應用希爾伯頻譜分析法，討論瞬時頻率。我們探討所接收 G.729 語音位元串流位元錯誤變化，希爾伯特黃時頻特徵變化之關聯性。我們所討論希爾伯特黃時頻特徵包括: (1)各個本質模態函數與殘餘函數平均值、最大值、最小值、標準差、能量、能量分佈比統計特徵；(2)各個瞬時頻率平均值、最大值、最小值、標準差；(3)各個本質模態函數在小於 65 Hz，65-175 Hz、175-494 Hz、 494-1319 Hz、1319-2093 Hz、2093-3952 Hz 和大於3952 Hz 各頻帶之能量和能量分佈百分比；(4)各個本質模態函數在各個時間間隔，在小於 65 Hz，65-175 Hz、175-494 Hz、 494-1319 Hz、1319-2093 Hz、2093-3952 Hz 和大於3952 Hz 各頻帶之時間能量和時間能量分佈百分比；(5)具有傳輸位元錯誤率之G.729 語音訊號各個本質模態函數和原始語音訊號各個本質模態函數間平均均方誤差；(6)具有傳輸位元錯誤率之G.729 語音訊號各個本質模態函數和原始語音訊號各個本質模態函數間皮爾森相關係數。從這個計畫，我們瞭解 G.729 壓縮參數和Matlab 水中聲波多輸入多輸出OFDM 多媒體通訊軟體平台傳輸位元錯誤率參數和語音訊號希爾伯特黃時頻特徵之關聯性。另一方面，我們在 B&K 我國代理商技術支援下，將應用水下麥克風和水聽器，建置G.729 水中聲學語音訊號通訊系統，探討不同傳輸距離參數和不同傳輸功率參數，所接收 G.729 語音訊號之希爾伯特黃時頻特徵關聯性。 The project explores using the G.729 voice compression algorithm for digital underwater acoustic audio communications based on Hilbert-Huang transform (HHT) time-frequency characteristics. We entered G.729 voice bit streams into the Matlab-based underwater acoustic multiple-input and multiple-output (MIMO) OFDM multimedia communications software platform developed in the project NSC99-2221-E-022-019 and generated G.729 underwater acoustic voice communication signal bit streams as the data output for decompression. A received and decompressed audio signal was disassembled into several intrinsic mode functions (IMFs) and a residual function using the empirical mode decomposition (EMD) method. We also applied Hilbert spectrum analysis to explore instantaneous frequency. We also explored the relationship between the change in the received G.729 voice bit streams because of transmission bit error rates and the change in the HHT time-frequency characteristics. Our discussion on the HHT time-frequency characteristics included (1) the statistical characteristics of the average, maximum, minimum, standard deviation, energy, and energy distribution ratio values generated from each IMF and residual function; (2) the average, maximum, minimum, and standard deviation values generated from each instantaneous frequency; (3) the energy and energy distribution percentage of each band for every IMF below 65 Hz, 65 to 175 Hz,175 to 494 Hz, 494 to 1319 Hz,1319 to 2093 Hz, and 2093 to 3952 Hz and over 3952 Hz; (4) the time energy and time-energy distribution percentage generated by each band for every IMF at time intervals below 65 Hz, 65 to 175 Hz,175 to 494 Hz, 494 to 1319 Hz,1319 to 2093 Hz, and 2093 to 3952 Hz and over 3952 Hz; (5) the mean square error between each IMF of G.729 audio signal with the transmission bit error rates and each IMF of the original audio signal; and (6) Pearson’s correlation coefficient between each IMF of G.729 voice bit streams with transmission bit error and each IMF of the original voice signal. This study has identified the correlation between G.729 compression parameters, the transmission bit error rate of parameters generated from the Matlab-based underwater acoustic MIMO OFDM multimedia communications software platform, and the G.729 voice signal of the HHT time-frequency characteristics. In addition, with the technical support of our domestic agent B&K, we employ underwater microphones and hydrophones to construct a G.729 underwater acoustic voice signal communication system. The system is used to investigate the received HHT time-frequency characteristics of G .729 voice signals to correlate the distance and transmission power parameters of various transmissions.

Keyword(s)

G. 729
水中聲學語音通訊訊號
希爾伯特黃
時頻特徵
Matlab水中聲波多輸入多輸出OFDM多媒體通訊軟體平台
傳輸位元錯誤率
平均均方誤差
皮爾森相關係數
水下麥克風和水聽器
G. 729
underwater acoustic voice communication signal
Hilbert-Huang transform
time-frequency characteristics
Matlab-based underwater acoustic MIMO OFDM multimedia communications software platform
transmission bit error rates

DSpace CRIS

A Hht-Based Time-Frequency Features Analysis for Underwater Acoustic G.729 Voice Signal-3

基本資料

Description