Innovative Sonar Design and Processing for Enhanced Resolution and Target Recognition

Some of the innovative sonar design and processing for enhanced resolution and target recognition projects include:

• Sonar Calibration
• Bathymetry from Klein 5410
• Multibeam Sonar on Klein 7180
• Time Synchronization

Related Publications for Sonar Capabilities

Sonar Calibration Facility

PRIMARY CONTACT: Barbara Kraft

We continue to make progress in the upgrades to our sonar calibration facility (funded in part by NSF), now one of the best of its kind in New England. The facility is now equipped with a rigid x, y positioning system, computer controlled transducer rotor (with resolution of 0.025 degree) and custom built data acquisition system. Barbara Kraft and Glenn McGillicuddy have reworked the software used for calibration. Measurements that can now be completed include transducer impedance (magnitude and phase) as a function of frequency, beam patterns (transmit and receive), open circuit voltage response (receive sensitivity), and transmit voltage response (transmit sensitivity). In addition, the A/D channel inputs have been optimized as a function of beam angle and the cross-correlation and RMS levels of the transmitted and received channels can be computed in real-time. Glenn is also working on the design of a new underwater pitch and roll stepper motor which will add new degrees of freedom to our calibration capability.

In the past year the calibration facility was used to better understand capabilities of several sonars including:

• Imagenix Delta T: The Imagenix Delta-T a new compact, low-power multibeam sonar that has been designed specifically for AUV deployment. The Delta-T is specified to form 120, 3 degree beams over a swath width of 120 degrees. Barbara Kraft and Glenn McGillicuddy performed calibration tests on the Imagenex DeltaT multibeam sonar at the UNH Acoustic Calibration Facility in order to verify the operational frequency, the transmit and receive beam patterns, the transmit pulse width and the source level at a range of system settings. Our hope was to integrate the Delta-T with the FETCH/DOERRI AUV (see discussion later in report) to explore the feasibility of collecting hydrographic quality bathymetric data and seafloor characterization data from an AUV. Calibrations of the Delta-T verified its operational frequency (260 kHz). Imagenex specifies the beam width to be 120° by 3° on both transmit and receive; however, the 3 dB beam width was measured to be approximately 90° while a beam width of 120° corresponded to 6 dB. The transmit beam pattern was measured repeatedly with an NUWC USRD E27 hydrophone to ensure that the it was aligned correctly with the DeltaT transmit transducer and any pitch and roll offsets due to the mounting hardware were resolved. The transmit beam pattern remained unchanged. To determine the receive element spacing and phase offsets, DeltaT native 837 files were recorded while a calibrated pulse was transmitted by the E27 hydrophone and received by the DeltaT in one degree increments while rotating the DeltaT from ±45°. DeltaT 837 files were also recorded for each setting of receive gain (1 dB steps from 0 to 20 dB) and display gain (20% increments from 0 to 100 %). All tests were repeated following repairs and upgrades to the DeltaT and the angular range of the receive tests was extended to ±80°. MatLab scripts were created to read the 837 files and process the raw receive element data for phase difference as a function of rotation angle as well as to simulate the quadrature sampling with digital mixing processes that may be applied by Imagenex during the DeltaT processing. Matlab code was also created (based on code from Christian de Moustier) to perform split aperture beam forming using two and three subarrays.
• ISSAP Probes: Kraft and McGillicuddy completed impedance measurements on the 40 and 65 kHz ISSAP probes. These probes are used in the ONR-sponsored In-Situ Sound Speed and Attenuation Probe, a device lowered to the seafloor to make in-situ measurements of seafloor acoustic properties.
• Env. Acous. Lab: Impedance measurements were made on five SensComp 40LT10 transmitters and twenty-five 40LTR10 receivers in support of the Environmental Acoustics Course lab exercises.
• EM3002/7125: Calder and Malik used the Acoustic Calibration Facility to calibrate the performance of two new Multibeam systems (the Kongsberg EM3002 and the Reson 8125) in order to better understand the long-term instrumental stability so as to help establish base-line uncertainty measurements for these systems. Included in these calibrations were measurements of transmit and receive beam patterns as well as long-term variance in performance at a range of instrumental settings.

Related Publications for Sonar Calibration

Bathymetry from Klein 5410

PRIMARY CONTACT: Christian de Moustier

With the successful completion of his Master's thesis, Jim Glynn, working with Christian de Moustier and Lloyd Huff has demonstrated that when sufficient care is applied to calibrating the Klein 5410 Sidescan sonar system, and with phasor averaging over three baselines, it is possible to obtain hydrographic-survey-grade bathymetry over swaths 150 m wide in 20 m of water depth or less. As a result of this work, Klein has re-wired the sonar arrays to access elements with a wider separation for more accurate phase measurements. The phasors are now obtained with elements spaced at 2.5, 4, and 6 wavelengths apart as compared to the earlier separation of 1.5, 2.5, and 4 wavelengths. The Matlab code developed by Glynn has been converted to C/C++ by Brian Locke so that it can be implemented in real-time. Additionally, the code has been modified to accommodate the new element spacing and make some corrections to the processing chain. A more reliable method of determining the actual element spacing has been devised, tested and implemented. This code; as well as suggested modifications to the array, were tested at sea on three separate cruises, two of these in direct collaboration with NOAA's Office of Coast Survey.

Brian Locke also designed and implemented modifications to the Generic Sensor Format (GSF) format to allow it to handle data from bathymetric sidescan sonars like the Klein 5410 (with separate port and starboard returns). The proposed changes have been submitted to SAIC who maintain the GSF format specifications and should be incorporated into the GSF specification in its next release.

In August, the 5410 with its rewired elements was used to collect data at the entrance to Portsmouth Harbor, NH in support of the thesis work of NOAA Corp officer Marc Moser on the R/V Coastal Surveyor under the supervision of Christian de Moustier and Lloyd Huff. For this cruise a "sled-bracket" was designed for the 5410. This sled-bracket has been requested for use on board both of the NOAA vessel Rude and Bay Hydrographer. The data collected on this cruise suffered from strong refraction-induced caustics, questionable CTD data, dynamic environmental conditions, and an inadequate PC for running the sonar and logging data (many survey lines suffered various instances of data loss because the PC could not keep up) and are still being evaluated.

In September, Jim Glynn and representatives from NOAA's Office of Coast Survey used the newly modified 5410 to resurvey the area south of Seavey Island in Portsmouth Harbor, New Hampshire.

Finally, in October, Jim Glynn and Brian Locke joined the NOAA ship Rude in Norfolk, VA to help the shipboard party setup the Klein 5410 and run a survey on the vessel's pole mount modified to accept the UNH "sled-bracket." This was a test of the first release of the C/C++ code, as well as a test of timing synchronization involving a timeserver and an IRIG-B time card in the sonar's topside unit. Further modifications to Brian's code have been made since this trip to make it more robust under various field conditions.

Based on this work, both Klein and NOAA's Office of Coast Survey have expressed an interest in finding a means of providing OCS with reliable instrumentation for a phase measuring bathymetric Sidescan. A meeting was held at Klein in Salem, NH to discuss possible avenues to go forward on this idea. It is not clear at this time the extent to which Klein is committed to such a system; however, eventually Klein did agree to re-vamp an existing Klein 5000 owned by OCS into a 5410. Consultations were provided to OCS during their negotiations with Klein that ultimately led to the re-vamping.

Multibeam Sonar on Klein 7180

PRIMARY CONTACT: Lloyd Huff

Lloyd Huff collected and evaluated data from the Multibeam subsystem which was mounted on the pre-prototype Klein 7180 Long-Range Side Scan Sonar (LRSSS) during its use on the Fairweather in the eastern Bering Sea (data collected in August 2006). These data have been subjected to intense scrutiny to study the idiosyncratic behavior of the measured depths. While funded from another NOAA grant (to Huff) this work is very relevant to the overall objectives and research goals of the Center. It is widely recognized that Multibeam bathymetry systems often have a specific cross-track pattern in depth measurement uncertainties stemming from the transition from amplitude detection of the seabed location to phase detection of the seabed. However, the cross-track pattern in depth measurement uncertainties in the pre-prototype Klein 7180 was quite different. Detailed investigation revealed that signals from another acoustic subsystem on the 7180 were being "folded" into the pass-band of the Multibeam sonar on the 7180. Since the data from the 7180 is only available after it has been subjected to several digital processing steps, it was necessary to develop a scheme of modeling/simulation to unravel the mystery. The modeling/simulation technique was the subject of an abstract submitted to OCEANS'07.

High-precision, High-accuracy Time Synchronization

PRIMARY CONTACT: Brian Calder

The ultimate accuracy achievable from a Multibeam survey can often be constrained by our ability to synchronize the time-stamps amongst the varied sensors (sonar, GPS, motion-sensor, etc.) associated with a survey. Brian Calder has been investigating the use of the IEEE-1588 'Precision Time Protocol' (PTP) as a solution for low-overhead time synchronization, primarily in survey systems (i.e., to allow local time-stamping at data generation as a way to eliminate latency issues in data capture). He has been able to demonstrate that on low-specification hardware (both computers - 533MHz Pentium III systems - and network - desktop workgroup 100bT Ethernet switches) the National Instruments PCI-1588 cards achieve synchronization and syntonization of clocks within approximate 100ns rms with zero host computer overhead, and low network overhead. Additionally he has demonstrated that a software implementation of the PTP can potentially achieve sub-millisecond accuracy when talking with a hardware master clock. The limiting accuracy is likely to be on the order of a few hundred microseconds, depending on computer speed and loading. The uncertainty in developing a timestamp from software, even using hardware oscillators, can be significantly higher than the hardware uncertainty. The estimate of this uncertainty is on the order of 10-20 microseconds depending on computer speed and loading.

Experiments were also done (with Andy McLeod) that demonstrated that an implementation of this approach over commercial wireless is limited in accuracy due to variable latency in the wireless switches, on the order of 1-5ms rms, with some spikes to 10ms. In support of this approach, Calder has developed code, termed the Software Grandmaster (SWGM) Algorithm, to synchronize, syntonize and absolute reference PTP time to a UTC master, in particular the 1PPS and ZDA messages from a GPS or IMU. The short-term accuracy of this system is typically 100-110ns rms from master to slave (slightly better on the master), and the long-term stability is essentially that of the GPS or IMU system itself. That is, SWGM-derived hardware times track UTC time absolutely within 100-110ns rms as long as the system remains in operation. The SWGM algorithm is robust to network packet loss up to ~60% and the a priori uncertainty estimated for timestamps by the SWGM algorithm match the true errors observed in the test environment. The self-timing of software latency is possible using processor register timing and careful control of process priority, but that process priority can significantly affect likely uncertainty in time-stamping (by an order of magnitude or more).

The value of this new timing synchronization approach was demonstrated this year when, under the supervision of lab manager, Andy McLeod, it was used to synchronize position, velocity and strain sensor data during tow-tank experiments that took place over a five week period. Over this period the timing synchronization between the motion controller and the data acquisition computer had an RMS error of approximately 620 microseconds over a wireless link.

Also in 2007, Calder demonstrated the value of this approach to NOAA through the integration and support of the SWGM software with a Reson 7-P Operating System, creating an experimental high-precision timing Reson 7125 Multibeam sonar system deployed on NOAA hardware. This system was demonstrated on the NOAA ship Bay Hydrographer in April 2007, confirming the performance in the field and providing the ability to measure latencies in other timing schemes.