Underwater Acoustic Barriers 2007 (UAB 07)

Transcription

1 Underwater Acoustic Barriers 2007 (UAB 07) NTNU - Trondhjem Biological Station- Outdoor Marine Basin, NTNU Sletvik Field Station NTNU Trondhjem Biological Station Standing acoustic field testing EC contract no Status: final Date: Sep 2007 Infrastructure Project Campaign Title NTNU - Trondhjem Biological Station- Outdoor Marine Basin, NTNU Sletvik Field Station Underwater Acoustic Barriers 2007 (UAB 07) HyIII-NTNU-5 NTNU Trondhjem Biological Station Standing acoustic field testing Lead Author Sergio M. Jesus sjesus@ualg.pt Contributors António Silva asilva@ualg.pt Date Campaign Start Date Final Completion 04/09/ /03/2010 Date Campaign End 15/09/2007

2 Contents: Heading: Contents: 1 Scientific aim and background 2 User-Project Achievements and difficulties encountered (max 250 words) 3 Highlights important research results (max 250 words) 4 Publications, reports from the project 5 Description 5.1 General description, including sketch 5.2 Definition of the coordinate systems used 5.3 Instruments used 5.4 Definition of time origin and instrument synchronisation 6 Definition and notation of the experimental parameters 6.1 Fixed parameters 6.2 Variable independent parameters 6.3 Derived parameters and relevant non-dimensional numbers 7 Description of the experimental campaign, list of experiments 8 Data processing 9 Organisation of data files 10 Remarks about the experimental campaign, problems and things to improve 1. Scientific aim and background: The global aim of the Underwater Acoustic Barrier 2007 (UAB 07) experiment was to demonstrate, in real conditions, some of the most recent technological and scientific developments in underwater acoustics. In particular experiments were conduced to demonstrate the application of an Environmental-based Equalizer (EbE) and the use of Orthogonal Frequency Division Multiplexing (OFDM) techniques for underwater communications (Part B1 and B2 respectively), and the usefulness of using Underwater Acoustic Barriers (UAB) for the protection of sea infrastructures (Part A). Moreover the experiments were used as engineering test (Part C) for innovative prototypes, developed by CINTAL, for underwater acoustic exploration. Despite the main scientific objective of the UAB experiment was comprised in Part A, Part B and C took place in first place, during the first week of September, since several hardware tests and adjustments were required to perform Part A (that took place in the second week of September). The data gathered during the experiments contributes to three main scientific objectives (Parts A, B1 and B2) and several technological objectives (Part C): Part A In this part of the experiment a setup composed of two interconnected vertical arrays: one transmit only array and one receive only array, was used to implement an underwater acoustic barrier for submerged intruder detection. It intends to show that using the time reversal principle, where the ocean is used as a spatial matched filter, signal energy can be simultaneously focused on each receive (only array hydrophone) and thus obtain the detector output by simple summation of the received energy over the array. This setup effectively configures a multistatic system with several transmitters and several receivers coherently processed both in time and space. Part B1

3 This part of the experiment comprises several underwater communication experiments with BPSK modulation schemes aiming at field testing an EbE for underwater communications named Frequency Shift passive Time-Reversal (FSpTR). The EbE aims at minimizing the MSE by taking in consideration the environmental properties that are varying during the data transmission. In its present implementation the FSpTR allows for the compensation of the source/receiver depth and source-receiver range variations. An experiment named "source-yoyo" was specifically designed to demonstrate with real data that the source depth variation makes a channel IR frequency shift and that the knowledge of such frequency shift can be used to improve the communications performance. For that propose two types of signals were transmitted: BPSK modulated signals at low, mean and high frequency, and chirp/lfm signals. The chirp signals are useful for channel impulse response and channel Doppler spread functions estimation which allows for a better understanding of the channel dynamics. Field calibration experiments with acoustic signals being transmitted in two different bands were performed for further study of IR characterization and channel capacity comparison in different bands. In addition a Multiple Input Multiple Output (MIMO) experiment was conducted with the use of the same signal being transmitted simultaneously in two sources at different depths. Part B2 This work addresses the problem of OFDM transmission in dispersive underwater channels where impulse responses lasting tens of miliseconds cannot be reliably handled by recently proposed methods due to limitations of channel estimation algorithms. The proposed approach relies on passive time reversal for multichannel combining of observed waveforms at an array of sensors prior to OFDM processing, which produces an equivalent channel with a shorter impulse response that can be handled much more easily. Part C Simultaneously with the scientific objectives the UAB 07 was used, also, as an engineering test for several technological equipments under development at CINTAL to support underwater acoustic research. Those devices comprise a set of electronic hardware assembled in three main units: (i) a remote stand-alone telemetry buoy that comprises an array of hydrophones and a PC104+ based telemetry unit with data storage and processing capability, named Acoustic Oceanographic Buoy (AOB) that is now in its final stage of development; (ii) a connected to shore base station that comprises a computer-based signal generator, a signal amplifier and an acoustic source; (iii) a transmit only array that comprises two acoustic sources and that operates shore connected. Moreover units (i) and (ii) are WLAN equipped for the acquired signals real time monitoring at the base-station side and GPS equipped for timing, synchronization and localization. During Part C of the experiment the use of Multiple Output Multiple Input (MIMO) system was also developed and tested with the main objective of being used in Part A of the experiment. The UAB experiment was carried out at Hydralab III - NTNU during the first two weeks of September During the first week (days 4 to 6) experiments were conducted in the Trondheim fjord to study and demonstrate point-to-point (P2P) underwater communication schemes based on EbE with Phase Shift Keying (PSK) modulation and Orthogonal Frequency-Division Multiplexing (OFDM) modulations. Those experiments took place at the NTNU biological station for shore operations and made use of the ship Gunnerus for offshore operations. During the second week (11 to 14 of September) the experiment took place in the Sletvik fjord to demonstrate the underwater AB concept for submarine intruder detection. That experiment took advantage of the Sletvik field station facilities and other facilities provided by NTNU personnel to meet the particularities of the experiment. Due to the complexity and the large number of experiments that were carried out during UAB 07 in this document only the Part B1 is described, Part B2 and A are described in separate documents: "

4 Preliminary analysis of UAB'07 multicarrier communications data" for Part B2 and "Underwater Acoustic Barriers Experiment UAB 07 Part A: the Hopavagen Bay" for Part A of the experiment. Those files can be found in Section 8 in "rep_uab07_b2.pdf" and "rep_uab07_a.pdf" respectively. 2. User-Project Achievements and difficulties encountered: To achieve Part B of the UAB experiments CINTAL personnel use several facilities: (i) for shore operations, a storage facility close to the NTNU-biological station to install their base station and the pier crane to place the acoustic sources at precise depths; and for offshore operations, the vessel Gunnerus to deploy and replace the AOB into appropriate locations, in addition the Gunnerus CTD was used to perform several CTD casts for sound speed profiles acquisitions. During this part of the experiment only small difficulties were encountered nevertheless they were readily solved by NTNU and CINTAL personnel. To achieve Part A of the UAB experiments CINTAL personnel make a visit to the Sletvik facility during the first week and identified the required facilities to carry on the experiments. It was identified that a close to the lagoon house and a small boat would be required. NTNU provide those facilities and in Part B of the experiment they were ready to be used. 3. Highlights important research results: Part B1 tested the source depth and array depth variation with the frequency shift (FSpTR equalizer) tracking channel impulse response variability. Results showed a source depth shift generating a frequency shift of the channel IRs partially compensating for lost performance caused by environmental variation. For a 0.5 m depth shift a frequency shift of 300 Hz occured, and a FSpTR equalizer improved 9.2% to 1.6% of wrong bits. The surface wave motion caused a frequency shift and a depth variation in the AOB hydrophones. Part B2 showed that the use of passive time reversal as a multichannel combining preprocessing step for impulse response shortening in OFDM receiver architectures was more reliable to multichannel combining performed after FFT processing and channel estimation at individual sensors, which failed due to the long impulse responses requireiing a large fraction of pilot tones for proper channel estimation. Time reversal stabilized the equivalent impulse response, reducing the impact of individual channel variations. Part A tested an acoustic barrier for submersed intruder detection. The hardware setup allowed for the implementation of a time-reversal (TR) detector using the actual acoustic channel to account for the matched-field receiver (the channel is stationary and a previous training phase is carried out with no target present). The non-channel based detector provided the lowest performance; the trivial detector achieved a slightly better performance, while the TR-based closely achieved the theoretical optimal performance. The system can be significantly improved by using more populated source and receiving arrays, and these results can be reproduced at sea in harbor like conditions. 4. Publications, reports from the project:

5 [1] Jesus S.M., A. Silva, C. Martins and F. Zabel (2008). Underwater Acoustic Barrier 2007 (UAB 07) Experimental Data Report. Rep. 02/08, SiPLAB Report, University of Algarve, February [2] Gomes J.P., A. Silva and S.M. Jesus (2008). Experimental assessment of time-reversed OFDM underwater communications. Proposed to Acoustics 08, Paris, June [3] Gomes J.P., A. Silva and S.M. Jesus (2008). OFDM Demodulation in Underwater Time- Reversed Shortened Channels. Oceans 08, Quebec, Canada, September. [4] S.M. JESUS and O. RODRÍGUEZ, ''A time-reversal suboptimal detector for underwater acoustic barriers'', OCEANS'08, Quebec, Canada, September, [5] A. SILVA, F. ZABEL, C. MARTINS, S. IJAZ and S.M. JESUS, "An environmental equalizer for underwater acoustic communications Tested at Hydralab III'', Hydralab III Joint User Meeting, Hannover (Germany), February, [6] S. IJAZ, A. SILVA, S.M. JESUS, "Compensating for source depth change and observing surface waves using underwater communication signals", (Submitted to) SensorComm 2010, Venice (Italy), July, Description: 5.1. Description: The UAB 07 experiment was carried in two different locations: at the Sletvik field station in the Hopavagen bay for Part A of the experiment and at the NTNU biological station in the Trondheim fjord for the Parts B1 and B2. Part C of the experiment, devoted to engineering tests, took place at both sites. In the following only Part B1 (and partially Part C) is described while Part A and B2 are described in separate documents "rep_uab07_a.pdf" and "rep_uab07_b2.pdf" ( see Section 8). Part B1 and C of the experiment in the Trondheim fjord Between 4 and 6 of September the experiments were carried out in the NTNU Biological station in Trondheim in Norway to perform Part A1 and A2 of the UAB 07 experiment. Photo 1a shows the storage house where the base station was installed and Photo 1b shows part of the basestation hardware that comprises two main units: one that generates and amplified the signals to be transmitted and another one that remotely monitories the AOB telemetry system.

6 Figure Storage house where the base station was installed (a) and signal generating and AOB monitoring equipment (b). The acoustic sources used to project the underwater acoustic signals were supported by a crane (fig 5.1.2a) close to shore. Low, medium and high frequency signals have been transmitted during the experiment; fig 5.1.2b shows the low frequency source. Figure Crane used to support the acoustic sources (a) and Low Frequency acoustic source Lubel 911 (b). The signals were acquired by the AOB acoustic telemetry system (fig 5.1.3a) that comprises a 64 m array of hydrophones 4 m spaced (fig 5.1.3b).

7 Figure Acoustic Oceanographic Buoy (AOB) prototype, developed by CINTAL, being prepared for deployment (a) and its 64 m array of 16 hydrophones (b). The AOB was deployed in the Bay of Trondheim (Figure 1a) in an environment characterized by a high down slope bathymetry with a water column depth of approximately 16 m at the source location and more than 100 m at the array location (Fig 5.1.4b).

8 Figure Trondheim bay map where the experiment took place (a) and an example of a transept between the source and the AOB during day 4 of September (b). To better understand the environment, in terms of acoustic channel, several chirp/lfm signals have been transmitted. Figure 2a shows the IR estimates obtained by the correlation of the chirp/lfm signal with the received data at the top ten hydrophones. There it is clear that the channel is characterized by two strong arrivals and then by a combination of unstructured multipaths. BPSK communication signals were transmitted to demonstrate the capability of the Environmental-based Equalizer to track source depth variations. By using the crane the source depth is changed at various known instants on time. This is called a YOYO experiment and is shown in fig. 2b where the time axis illustrates the time starting from 12:40 pm.

9 Figure Channel IR estimate obtained by pulse compression of the transmitted chirp/lfm signal example took during the first day of operation - (a) and source depth Yoyo operation - example starting at 12:40 GMT of the first day of operation - (b). The FSpTR block diagram used to demodulate and equalize the transmitted BPSK signals is shown in figure 3a and an example of the attained results is shown in figure 3b. At the 6th second of figure 3b a source depth shift from 4 to 4.5 m occur and it can be see that a strong performance loss occurs for the ptr (the same as FSpTR bur without frequency shift) demodulating system, however the FSpTR partially compensates for that performance loss by maintaining the mean squared error lower than -2 db. Figure Left side of (a) shows the probe and data signals underwater propagation; Right side shows a block diagram of FSpTR equalizer: (i) filtering of hydrophone received data with time-reversed FS IR estimates, (ii) addition of filtered signals for each FS, (iii) selection of the FS signal with maximum power, (iv) down-sampling to the symbol rate and (v) estimate of transmitted symbols (a). Mean squared error of the ptr and FSpTR communication systems, comparison for 10 hydrophones with a source depth shift at 6 sec (data set 1) (b). In addition to the acoustic signals the experiment comprises the measurement of environmental signals as source and array depth variations, temperature, and sound speed profile. The last was acquired by CTD casts performed on board of Gunnerus (Photo 4). Photo 4b shows the sound speed profile extracted from the CTD cast during 4 of September.

10 Figure Gunnerus crew operating the CTD (a) for sound speed profiles acquisition (b). The Part C of the experiment comprises several technological equipment tests. Those tests comprise electrical and mechanical robustness tests of the prototypes developed by CINTAL.

11 Photo 5a shows the vessel Gunnerus that was used to the offshore operations and 5b shows the AOB being towed by Gunnerus wile in operation. Figure Source depth Yoyo operation, example stating at 12:40 GMT (a) and AOB being towed by Gunnerus to test its robustness wile in operation (b). During the experiment we receive visitors from SINTEF. Figure Visitors: SINTEF personnel (a) and Jense Hoven and Fred waiting for the Gunnerus to track (b). Instruments used: During the experiment the following instruments have been used: - Acoustic Oceanographic Buoy (AOB) telemetry system that is able to acquire underwater acoustic signals in the 0.1 to 16 khz band with a sampling frequency of 50 khz. The AOB is a free drifting buoy that comprises 16 hydrophones 4 m spaced in a 64 m array, a GPS system for self location and timing synchronization and a WLAN for remote monitoring. The acoustic and

12 GPS data are stored in VLA files that can be read with the Matlab file readlocapassdata.m that can be found in Appendix A of [1]. - HOBO datalogger, that acquires the absolute pressure in [mbar] and temperature in [ºC]. During the experiment it was used to monitories the AOB array depth and the acoustic source depth. - CTD standard devise, that allows to measure the pressure/depth, temperature, salinity and other parameters along the water column. For experiment purposes the acquired data was manly used to derive the water column sound speed profile (ssp). - Acoustic source/projector Lubell 911 that is able to transmit underwater acoustic signals in the 0.2 to 8 khz band. - Acoustic source/projector Lubell LL916 that is able to transmit underwater acoustic signals in the 1 to 20 khz band Definition of the coordinate systems used: 5.3. Instruments used: 5.4. Definition of time origin and instrument synchronisation: All the measurements have been GPS synchronized in GMT time. 6. Definition and notation of the experimental parameters: 6.1. Fixed parameters: Acoustic signals parameters: The structure of the transmitted acoustic signals during Part A1 experiment comprises a set of chirps/lfm repeated N times (for channel impulse response evaluation), followed by BPSK modulated stream with 1 second duration repeated M times as shown in the Table 6.1. Table 6.1. Acoustic signals format At the beginning of each BPSK 1 second signal an M-sequence of 127 symbols was transmitted for synchronization and for channel estimate during the data transmission. The pulse shape of the transmitted BPSK symbols is a forth root raised-cosine pulse. Five different types of signals were transmitted during the experiment. The BPSK part of the signals can be characterized by its symbol rate, carrier frequency, band of frequencies used and number of repetitions. The Chirp/lfm part of the signals is characterized by its band, duration and number of repetitions. Table shows the parameters of each one of the signals. The signals were generated in Matlab 5.3 with a sampling frequency of 50kHz, the signals were then converted to a text file to suit the base-station requirements. Table shows the MATfile names that contains detailed information about the transmitted signals. Table Acoustic signals characteristics

13 Table Mat-files with detailed information about the transmitted signals. On the AOB side the received signals were acquired with a sampling frequency of 50kHz, and stored in VLA format. The VLA files-names have information on the time when they were acquired in the following format "AOB vla" where: "AOB2" is telemetry equipment AOB version 2, "247" is the Julian day and "124848" is the GMT time (12 hours, 48 minutes and 48 seconds). Due to the high sampling frequency of the signals a huge amount of data was acquired during the experiment and only a small set of that can be provided in Section 8 (at the ReceivedEx directory). In Section 8 it can, also, be found an m-file "ReadLOCAPASSData.m" that allows to open the vla-files in Matlab. Those who are interested in processing the data can contact Antonio Silva (asilva@ualg.pt) and/or Fred Zabel (fredz@wireless.com.pt) to recieving the data or to clarify any lack of clarity in this report. Environmental/non-acoustic parameters: Table shows the non-acoustic parameters that were measured during the experiments. The objective of the CTD measurements are manly to monitories the watercolumn sound speed profile, the HOBO measurements were used to monitories the array and source depths and the GPS data were used for time synchronization and location. Table Environmental parameters 6.2. Variable independent parameters: Notation Name Unit Definition Remarks

14 Table Derived parameters and relevant non-dimensional numbers: Notation Name Unit Definition Remarks Table Description of the experimental campaign, list of experiments: Experiment Name Experiment Date Remarks 7.1. Day 4 experiments and measurements: On day 4 of September (247 Julian days) the experiments were made to test the FSpTR equalizer with and without source-yoyo. The time schedule of the experiments is described in the table Table Day 4 experiments and time schedule A total of 5 CTD casts were done on board of Gunnerus at the following time and location: Table CTD casts during day 4 During the experiments the HOBO datalogger was placed close to the array middle. Those data were stored in the "uab txt" file (see Section 8). The AOB movement during this day can be found in the "uab07_gps log" file (see Section 8) Day 5 experiments and measurements: During day 5 of September (248 Julian days) field calibration experiments with and without sourceyoyo were performed with the following time schedule:

15 Table Day 5 experiments, time schedule A total of 12 CTD casts were done on board of Gunnerus at the following time and location: Table CTD casts during day 5 During the experiments the HOBO datalogger was placed close to the array middle. Those data were stored in the "uab txt" file (see Section 8). The AOB movement during this day can be found in the "uab07_gps log" file (see Section 8) Day 6 experiments and measurements: Day 6 of September (249 Julian days) was devoted to engineering tests and MIMO data communications those experiments have been performed with the following time schedule: Table Day 6 experiments, time schedule

16 A total of 8 CTD casts were done on board of Gunnerus at the following time and locations: Table CTD casts during day 6 During the experiments the HOBO datalogger was placed on the top edge of the cage of the top transducer and the bottom transducer was placed 0.5m down. Those data were stored in the "uab txt" file (see Section 8). The AOB movement during this day can be found in the "uab07_gps log" file (see Section 8). 8. Data processing: 9. Organisation of data files:

17

18 10. Remarks about the experimental campaign, problems and things to improve: The Hydralab III, gave to CINTAL the opportunity of experiment in real conditions a set of equipment that was in an initial stage of development as is the case of the base station and in a final stage of development as is the case of the AOB prototype. Moreover, it allowed the acquisition of acoustic signals to demonstrate: the feasibility of Underwater Acoustic Barriers for intruder detection, the possibility of using Environmental-based Equalizers for underwater communications and the usefulness of using the passive Time-Reversal concept for shortening pulses in OFDM communications. The CINTAL team found at Hydralab III NTNU quite nice facilities for the performance of underwater acoustic experiments. Among those it should be mentioned the shore facilities at the NTNU Trondheim Biological Station and the offshore facilities provided by the vessel Gunnerus. It should be noted that when those facilities were not available as is the case of a shore house and a boat at the Sletvik Field Station (at the Hopavagen Bay) they have been readily provided by the NTNU personnel and the experiments run easily. Window size: x Viewport size: x