Feature Articles—November 2009 Issue
An Improved Diver Detection Method For Shallow-Water Port AreasX-Type Diver Detection Sonar Adapts to Nonstationary Sea Environments for Improved Detection of Small Targets
By Dr. Victor D. Svet
Head
Research and Development
Donald Sandilands
Product Development Manager
and
Manel Monteiro
International Marketing
and Sales Director
Oceanscan Ltd.
Aberdeen, Scotland
Diver detection in shallow-water areas is characterized by multipath propagation, high bottom and surface reverberation, and a wide range of noise sources. These are all factors that strongly complicate detection, tracking and target classification in the limited time that is usually available for response.
An improved method of signal processing, based on experimental and simulated results, has shown that Oceanscan Ltd.’s X-Type diver detection sonar (DDS) can improve performance through the application of more advanced signal processing.
Surface and bottom reverberation is the basic noise found when using active sonar to detect small targets in shallow-water areas. High-angle-resolution sonar arrays combined with correlation compression of wideband echo signals can decrease the influence of reverberation significantly, and these technologies are widely used in a DDS.
Sources of Noise in Seaports
The nature of seaport noise sources has various influences on the performance of a DDS.
For example, coastal noise is connected with the work of numerous coastal mechanisms where a vibrating field is radiated into the water as acoustic waves. Accepting that the spectra of such noises are low frequency and, as a rule, lie outside of the working range of frequencies used for detection, their intensity can still be high enough to have an influence. It is necessary to also consider that the majority of coastal mechanisms are devices with rotating elements, and their noise can contain many high-frequency, quasiharmonic components.
Ship noise is created by vessels moving in the port. It is necessary to use well-known ship noise dependencies with great care, because estimations of ship noise levels are made in the far-field zone. This condition is usually not satisfied for the zone of operation of a DDS in a port. The effective spectrum of ship noise is below the working frequency band of a DDS, but due to the very high noise levels, part of the noise energy can be injected into the receiving array. In addition, the majority of vessels can also create high-frequency cavitation noise from propeller action.
Wakes are another source of noise that can have a serious influence on the operation of a DDS, especially if the vessel crosses a sector of view of the system at a certain angle and over a range of detection distances. Powerful turbulent hydrodynamic streams, appearing between the targets and receiving array, create an intense layer of phase inhomogeneities. A significant part of the transmitting signal dissipates in these inhomogeneities, and the echo signal from the target is not detected. It is important to note that powerful wakes can persist for some time in the water column.
Active acoustic signals generated by different hydroacoustic ship systems also create very strong interference. Practically all vessels are equipped with an echo sounder or a similar hydroacoustic system that operates constantly. The frequency band of such systems is very wide and sometimes very close to the frequency band of a DDS. Accepting that the basic energy of an echo sounder is distributed in the vertical plane, the inevitable side lobes of their directivity patterns in transmitting mode, when combined with the shallow depth, can create significant interference in a DDS’ receiving arrays. Because echo sounders’ pulse signals are not synchronized with the transmitting pulses of a DDS, they can create strong false marks in the resolution elements of the basic sonar picture.
Breakwater Installation Noise
For a DDS installed on a wall or breakwater, there is another specific noise source generated by breaking waves. From general data about sea state noise in the frequency band 30 to 60 kilohertz, it can be assumed that only a sea state greater than three will result in noise greater than the thermal noise of the electronics.
However, waves breaking on a concrete wall can generate acoustic noise from resonant air bubbles concentrated in the foam at the surface water layer as well as the noise of physical wave impact. This can increase the level of general noise around 100 kilohertz.
Finally, there is one more source of noise to consider on a breakwater: pseudo sound noise. Normally, this noise has no influence if the sonar is installed on the seabed. But pseudo-noise can be important for breakwater installations because this is a zone that generates surface waves and turbulent water flows. Actual pressure fluctuations can be very large at frequencies below 10 to 20 hertz. Although it would seem that such low-frequency noise should not create problems for a DDS working at a frequency of 100 kilohertz, the problem is that even such low-frequency variations of ambient pressure can generate large voltage spikes at the inputs to the array preamplifiers.
As a result of the influence of all these specified multipath propagation and noise factors, the received echo signals are characterized by strong space-time invariance and amplitude-phase fluctuations.
New Approach
A new approach has been developed based on exploiting the difference in time coherence between the useful echo signals and the nonstationary noise signals, including reverberation. Generally, it means that the detection threshold cannot be a constant while applying the detection procedures, but must be variable. This procedure has to be adaptive due to the nonstationary environment.
The detection threshold varies in different resolution elements with angle and distance, and it changes after several sets of the active probing sonar signal according to some criteria. For example, in more nonstationary environments, the threshold can be updated every five to six seconds. Generally the update time for the threshold depends on local sea conditions and the intensity of ship navigation in the port area. Adaptive threshold processing allows the stationary noise background, static targets and random, nonstationary echo signals to be effectively suppressed.
The result is that the DDS can confidently detect echo signals from small moving targets with significantly improved spatial resolution.
Sea Trials of Detection Procedures
Some examples of these adaptive threshold detection procedures demonstrate their effectiveness in different environments. Sea trials were completed in the North Sea (Scotland), the Indian Ocean (Sri Lanka) and the Yellow Sea (China).
The trials were carried out using the X-Type DDS designed by Oceanscan Ltd. This sonar has phased arrays and operates at 100 kilohertz (bandwidth 30 kilohertz) using complex modulated signals. It has a horizontal angle resolution of one degree and a source level of 217 decibels. The X-Type can be tripod-mounted and installed on the seabed or bracket-mounted and installed on a wall or breakwater.
In the Yellow Sea trials, using a diver with open-cycle breathing, it was clearly seen that standard threshold procedures—in which the value of the threshold is constant for all resolution elements—could not reliably recognize a target. However, it was demonstrated that consecutive adaptations of the detection threshold and automatic estimation of the optimal time series allowed confident detection of the target at large distances and suppression of stationary and nonstationary noise.
Other trials demonstrated the X-Type’s capability to use adaptive processing to detect two small targets close together in a nonstationary environment with high surface reverberation. These trials were carried out in the North Sea (Peterhead Harbour). In these experiments, an air bubble stream was towed from a small boat and used as a diver simulator. The air stream was modulated with a valve to simulate breathing. Adjusting the air pressure created bubble streams of different intensities that were equivalent to varying the equivalent target strength from -15 decibels up to -25 decibels.
The detection of a small boat as a surface target at large distances is more difficult, but can be accomplished with the X-Type. The small keel draft means the equivalent radius of the towboat is very small—comparable to, or even less than, the equivalent radius of a diver. Also, it is necessary to consider that such a boat, moving with a speed of less than one knot, rolls continuously due to wave action from the sea surface. The average distance between the diver simulator and towboat was about 12 to 15 meters. Due to underwater currents and wind angle direction, the course or position of the bubble stream and the boat could be slightly different.
Conclusions
During numerous trials in different oceans, the X-Type DDS has confirmed its ability to detect weak targets of small dimensions in shallow-water nonstationary environments. The methods developed for adaptive threshold signal processing have ensured reliable detection and tracking of many small moving targets at distances of 600 meters and more
Dr. Victor D. Svet worked for more than 37 years at the Institute of Acoustics, Moscow, as head of laboratory signal processing and imaging. He is the author of more than 160 papers and 20 patents. In 2005, he joined Oceanscan Ltd., and is currently head of research and development.
Donald Sandilands has worked for more than 20 years in the marine industry for various equipment manufacturers in engineering and engineering management positions. He joined Oceanscan Ltd. in 2005 and is currently product development manager.
Manel Monteiro has worked for more than 20 years in the marine industry for various equipment manufacturers in marketing and engineering positions and has worked on the development of the Diablo and Demon remotely operated vehicles as well as the Curvetech propulsion system. He is currently Oceanscan’s international marketing and sales director.
Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.