ABSTRACT

TitleAbstract
Compensation of Multipath Fading in Underwater Spread-Spectrum Communication Systems
   Techniques for achieving reliable communications over underwater acoustic channels are investigated in this work. It is well known that, due to multipath and Doppler effects, an underwater acoustic communication channel may exhibit rapidly fading characteristics [1]. This makes building a reliable underwater communication system a challenging task. Conventionally, a fading channel can often be compensated by using an adaptive equalizer. However, for an underwater acoustic channel, conventional equalizers may fail to properly track the rapidly changing channel characteristics. In this work, we consider implementation of an underwater spread spectrum communication system [2]. The spread spectrum technique was investigated originally for military communications and is widely used in many modern communication systems. Due to its relative immunity to interference and multipath fading, the spread spectrum technique is suitable for the underwater acoustic communication channel and thus is adopted in our system. The modulation technique used in our system is the pi/4-QPSK. Since demodulation of the pi/4-QPSK can be performed noncoherently, carrier phase recovery is not required and thus a relatively low complexity receiver can be achieved. Although neglecting the carrier phase deviation is justified when the pi/4-QPSK is used, the carrier frequency offset, which may degrade the overall system performance, can not be ignored. To eliminate the carrier frequency offset, an automatic frequency control (AFC) technique is used to control the output frequency of the receiver local oscillator. In order to compensate those channel impairments owing to multipath and Doppler effects, an effective adaptive equalizer is required. In this work, a transform-domain approximate RLS (TARLS) blind equalizer, whose operation requires no training sequences, is proposed to compensate the impairments in the underwater acoustic channel. Blind equalizers have been found useful in some mobile communication systems where training causes too much overhead. However, most conventional blind equalizers are based on the LMS algorithm and thus converge very slowly when the input signal is highly correlated. This prevents conventional LMS blind equalizers from being used in underwater acoustic channels. Compared to conventional LMS blind equalizers [3], the TARLS blind equalizer has comparable computational complexity but significantly better convergence and tracking performance. Simulation study indicates that the TARLS blind equalizer has good numerical stability in a finite-precision environment and thus is ready for digital implementation. This makes the TARLS blind equalizer a good choice for our system. The entire system is implemented on a DSP module with a TI TMS320C44 chip. The targeted data rate of this system is 10 k bits/sec. The advantages of the proposed system over its conventional counterpart are demonstrated by using a simulated underwater communication channel.
   This work was supported in part by the National Science Council under
   Contracts NSC 86-2611-E-019-018 and NSC 86-2611-E-019-020.
   [1] J. Catipovic, "Performance limitations in underwater acoustic telemetry," IEEE J. Oceanic Eng., vol.OE-15, pp. 205-216, July 1990.[2] R. L. Peterson, R. E. Ziemer, and D. E. Borth, Introduction to Spread Spectrum Communications, Prentice-Hall, 1995. [3] A. Benveniste and M. Goursat, "Blind equalizer," IEEE Trans. Commun., vol. COM-32, no. 8, pp. 871-883, Aug. 1984.

VENUS Project - Submarine Cable Recovery System
   Deep seafloor observatories utilizing submarine cables have been very attractive for oceanographic measurement. In Japan, developments of observatories are accelerated by the demand to construct seismic monitoring network and retirement of coaxial telecommunication submarine cables. A submarine cable laying between Okinawa and Guam, was terminated its twenty years services in 1994 and handed over to a national university for scientific re-utilization. Universities and several research institutes planned and carried out a project to reuse this cable under the funds from Science and Technology Agency of Japan. This is VENUS (Versatile Eco-Monitoring Network by Undersea-Cable System) project. The observatory of this project will be a first complex observatory which is installed to retired cable. Installation of the observatory requires three procedures which are cable amputation, cable recovery, and reinstallation. For VENUS project, the cable reinstalling operation needs precise cable position control to put the observatory to target area. It is very difficult to control the position of observatory using conventional cable recovery and installation method by grapnel anchor. Japan Marine Science and Technology Center (JAMSTEC) develops a cable recovery system to realize the precise cable position control. This system consists of a cable cutter, cable grippers and a hydraulic pump unit. The hydraulic pump unit which is electrically powered from outside, generates hydraulic pressure to actuate the cable cutter and cable gripper. This system has an ability to amputate the cable at the deep seafloor and pick out to the surface. Cable recovery system was designed and manufactured to fit for manned submersible 'SHINKAI6500' because of the VENUS observation point plot on map has 4200 meter depth. Pressure tests, amputation tests and strain tests are given to confirm the performance of this cable recovery system. The cable recovery operation for the target submarine cable will be conduct in fiscal year of 1997 at the VENUS first observation site in NANSEISHOTO-trench.

Coaxial Cable Information Transmit System of 6000m Deep Towing System
    In this paper, the coaxial cable, used in the transmission of underwater television images and data, is designed for 6000m deep towing observational system. The towing system is used in the exploring of underwater manganese nodule in Pacific Ocean.
    It's difficult to transmit underwater television images and data long distance through coaxial cable without relay amplifiers for deep water ROV. These problems lie in the fact that, the disturbance between the power energy and signals in one line, and the high frequency decrease in 9000m coaxial cable without amplifiers. So, two problems must be solved. The first one is to consider the width of frequency band of informational transmission. The second one is to adapt suitable multiplexing transmission technique. These two problems is solved very well in this system.

Study on the Dynamic Performance of the Cable Used in 6000m Deep Towing System
    The seabed is abundant with various minerals, such as manganese. Deep towing system can be used in exploring and exploiting manganese which deposit in the seabed. Manganese usually deposit in the seabed where the depth is over 5000m. On the other hand, the depth of 6000 has covered over 96 percent seabed area. So, this system limits its work depth to 6000m. Up to now, the system mentioned here had got a great success in the Pacific Ocean for two trials.
    In such a system, cable is the medium which links the mother ship and the towed body. It has the ability of conducting power and transmitting signals. Meanwhile, it's an important manner to balance the body in water. In order to undertake all these tasks, the cable usually adopts optical cable or coaxial cable with good tensile strength.
    In this paper, deep towing system is studied based on dynamic analysis. And the main idea is to study strain change in cable when the towed body responses to draw-in and let-out cable, as well as to heaving. The details are as follows:
   1. Building up the mathematical model for deep towing system
    Deep towing system consists of three parts, such as mother ship, cable and towed body. In this paper, we assume that the ship moves straight ahead, and the influence of wind and sea is neglectable, the whole system works in vertical plane only. Then a differential equation on the cable is built up. The equation is of four variables: cable tension, tangential and normal speed, local angle. A finite difference approximation has been used in solution. A rather new boundary condition and initial condition are presented.
   2. The dynamic performance of the cable
    As for the dynamic performance of the cable, the strain in both ends is particularly worth studying. And the length, the weight and the shape of the cross section the cable adopted, do a great deal to the performance. So, we made a few examples of these, and dug into the further reasons. Besides, studying on the strain change in cable is emphasized here when the towed body responses to draw-in and let-out cable, as well as to heaving. Some interesting phenomena can be observed.
    The study is valuable in guiding the selection in cable design. Also, the research of dynamic motion of the system can be used in guiding the operation.

A New Approach to Real Time Measurements at the Izu-Bonin Trench Using TPC-1, Demission Telecommunication Cable
    The present real-time seismic network in the global scale is not enough to get high resolution images for Earth's structure due to lack of seismic stations in the oceanic region. In earthquake prediction study, it is extremely important to monitor earthquakes at trenches by real time telemetry because of necessity on rapid analysis of seismic activities. Most of ocean bottom seismic observations have been done by self-contained storage OBSs (Ocean Bottom Seismometers). One of the best ways to built real time monitoring systems in the ocean is to use submarine cables. Brand new submarine cables, however, are very expensive if proposed locations are far distance from land. In order to overcome financial difficulty, a new approach to use demission submarine cables is proposed. The TPC-1 (Trans Pacific Cable 1) Guam-Ninomiya section was built in 1964 and terminated her commercial use. The Earthquake Research Institute in Japan and IRIS in US jointly obtained the TPC-1 Guam-Ninomiya to scientific purposes in 1990. The length of the cable is 2659km. The GeO-TOC cable (former TPC-1) uses 1-inch coaxial cables and 72 vacuum tube repeaters. A newly built OBS contains three axes accelerometers, a quartz thermometer, a quartz, pressure sensor, a hydrophone and thermometers to measure board temperature and accelerometer temperature. The 'IZU' OBS was deployed at the Izu-Bonin Trench slope (31deg. 24.62'N, 140deg. 54.33'E, 4200m deep) by splicing the GeO-TOC on January 1997 and the data from instruments have been transmitted to the ERI in Tokyo. The DC power used by instruments is taken from cable power supply that is supplied as 4170V and 370mA constant current from the Guam shore station. The output signals of one component seismometer is digitized by 24 bit and sent to shore by bit-rate of 9600 bps through a 4-kHz frequency carrier. Instrumental data except the hydrophone are transmitted to the Ninomiya shore station by digital form. The hydrophone data are sent to shore by analogy form and then digitized at shore. The whole data are re-transmitted to Tokyo by commercial telephone line. Seismometer gimbal, amplifier gains, CPU reset, modem gains inside of the OBS can be controlled from the laboratory terminal. The VENUS project in Japan and the H2O project in US will do the next steps for demission submarine cable use.