Zhu Weiqing

Underwater acoustic communication system of AUV

The underwater acoustic communication system (UACS) is an important part of an AUV. We developed an underwater acoustic communication system for a AUV: IASS-1, which is based on both MFSK and MPSK. Generally a AUV sends a large amount of data to the host vessel, and MPSK modulation is adopted. The receiver on the host vessel adopts space diversity (SD), a decision feedback equalizer (DFE), and fast self-optimized LMS phase estimator (FOLMSPE) techniques jointly. The fast self-optimized LMS (FOLMS) algorithm is implemented in the DFE. A short vertical hydrophone array is suspended or towed on the host vessel for signal reception. The separation between each hydrophone is far larger than the wavelength. The MFSK modulation and convolution code are adopted when the host vessel sends a message to the AUV. The receiver on AUV adopts the Viterbi algorithm. This well-known mechanism, using MFSK and convolution code, performs well when the transmission rate is low. Compared with MPSK modulation, it has the advantage of being very resistant to the most difficult multipath propagation, and is less sensitive to the vessel scattered wave. IAAS-1 was tested in a lake and was successfully operated at ranges over 100-4000 m (constrained by the test area). The symbol rate for QPSK is 2.5 ks/s and 5.0 ks/s. The bit error rate (BER) is between 10/sup -4/ and 10/sup -5/.

Source number estimation using eigenspace in direction of arrival (DOA) estimate

A source number estimation using eigenspace is presented. It projects estimated covariance matrix of array signal into signal eigen subspace and noise eigen subspace respectively. Using the orthogonality between signal eigen subspace and noise eigen subspace, it is easy to differentiate the contribution of signal and noise by using the criterion value, which is the magnitude of projection. Like the direction of arrival (DOA) estimate algorithm, the estimation uses the eigenvalue decomposition of covariance matrix with M×M order (M is the number of elements). Hence much computational burden can be saved. To reduce more computational burden, the estimation can be realized by the decomposition in real-valued space. Computer simulation demonstrates the distribution of criterion value and the performance on the condition of signal sources with equal power, with unequal power and space correlative color noise environment. The estimation was also tested with the sonar data. It is show that this estimation has good performances.

Signal Processing for High Speed Underwater Acoustic Transmission of Image (March 2007)

In this paper, a signal processing method for high-speed underwater acoustic transmission of image is presented. It has two parts. Part 1 is underwater acoustic coherent communication signal processing. Part 1 includes 3 technical points. 1. Doppler shift compensation. Chirp signals are inserted between data packages. A correlation process between two copy correlation results gives more accurate estimation of the mean Doppler shift. Resample the data to compensate it. In feedback adaptive equalizer an adaptive phase compensator with fast self-optimized least mean square (FOLMS) adaptation algorithm is utilized resulting in better motion tolerance than compensators with 2nd order PLL algorithm. The performance of the combination of mean Doppler shift compensation and adaptive phase compensator is quite good. 2. A combiner is used in advance of equalizer. Both combiner and feedback adaptive equalizer are based on FOLMS adaptation algorithm and have reduced computation complexity and good performance. 3. Cascaded equalizer and Turbo-TCM decoder and the iteration algorithm. A new bit-symbol converter based on Soft Output Viterbi Algorithm (SOVA) is studied. Comparing with the traditional decision, coding and mapping algorithm, new converter can reduce BER by nearly 2 orders. Part 2 is a robust image compression algorithm. In high-speed underwater acoustic image transmission, there are always some error bits. Ordinary image compression algorithms, for example, JPEG and JPEG2000, are sensitive to error bits. One error bit may destroy whole image. So they are not suitable for high-speed underwater acoustic image transmission. Based on digital wavelet transform and fixed length coding, a robust compression algorithm for optical and acoustic image is studied. The algorithm includes 4 technical points. 1. Utilizes CDF9/7 wavelet base to transform the images. 2. Analysis the energy distribution of the subbands coefficients. A suitable transformation layer number is 3. 3. Apply different quantization steps to different subbands in accordance with their energy distribution. 4. Fixed length coding to prevent error propagation. The results show the algorithm achieves a balance among image quality, compression rate, and most important, robustness to bit error. The data rate of compressed true color optical image and gray scale acoustic image are 0.80 bit/pixel and 0.85 bit/pixel respectively. Image quality remains good when BER is lower than 10-3. There are some small dirty points when BER rise to 10-2. Based on the signal processing techniques described above, an underwater acoustic communication system is built. Its band is (7.5~12.5) kHz. Its receiving array is an equal spaced 8 elements linear array. Each element is semi-spherical directivity. QPSK and 8PSK modulation and iteration algorithm for cascaded equalizer and Turbo-TCM decoder based on Soft Output Viterbi Algorithm are used. The system is tested in Qiandao Lake, a 20-50 meters deep reservoir with lots of submerged hills. Low BER is achieved in 5.5 km rang when data rate is 10 kbps. One 256times256 true color image or gray scale image can be transmitted in 7 second. The product of its communication distance and data rate is 55 km*kbps. Without the limitation of the lake area, the product is estimated to reach 80~100 km*kbps.

Theory and Signal Processing of Acoustic Correlation Techniques for Velocity Measurement of Vessel

The theoretical model and signal processing of correlation techniques to estimate the velocity of a vessel relative to the bottom are discussed. When the random component of velocity is considered, the sonar space-time correlation function of bottom reverberations within Fraunhofer zone is derived. The function, which is composed of zero order, first order and second order Bessel function, is the theoretical model of acoustical correlation techniques. By the model, “wave invariance” can be well described. Localized least mean squares (LLMS) criterion is put forward for velocity measurement. Sequential quadratic programming (SQP) method is adopted as the optimization method. A prototype correlation velocity log (CVL) underwent several deep sea trials, the results show that theoretical model approximately coincides with experimental data. Square loop tracklines of the ship in 3560 m deep area were adopted to examine the performance of CVL. Varying velocity cruise test and drifting test were also carried out. The CVL performed well in these tests. A least squares linear regression shows a significant correlation between the speeds of CVL and those of GPS. The rms error of CVL is less than (1.4%v+3) cm/s, where v is the velocity of vessel.

Design and Test of 70 kHz Correlation Velocity Log

A 70 kHz prototype correlation velocity log (CVL) has recently been developed. The hardware design and signal processing were discussed. Localized least mean squares (LLMS) was proposed as the criterion of velocity estimation. The CVL underwent a sea trial in the north of the South China Sea in August 2007. The CVL was calibrated in the beginning of the sea trial. Then square loop trackline test, varying speed cruise test and drifting test are carried out to examine the performance of the CVL. The uncertainty statistics from the sea trial showed that the rms error of the CVL was less than (1.2% v+3) cm/s.

Acoustic estimation of suspended sediment concentration

In this paper, the acoustic estimation of suspended sediment concentration is discussed and two estimation methods of suspended sediment concentration are presented. The first method is curve fitting method, in which, according to the acoustic backscattering theory we assume that the fitting factor K1 (r) between the concentration M(r) obtained by acoustic observation and the concentration M0( r) obtained by sampling water is a high order power function of distancer. Using least-square algorithm, we can determine the coefficients of the high order power function by minimizing the difference betweenM( r) and M0( r) in the whole water profile. To the absorption coefficient of sound due to the suspension in water we do not give constraint in the first method. The second method is recursive fitting method, in which we take M0( r) as the conditions of initialization and decision and give rational constraints to some parameters. The recursive process is stable. We analyzed the two methods with a lot of experimental data. The analytical results show that the estimate error of the first method is less than that of the second method and the latter can not only estimate the concentration of suspended sediment but also give the absorption coefficient of sound. Good results have been obtained with the two methods.