Home

Hydrophone equipped mote

1. start public void startSW2WatchThread new Thread public void run while true sw2 waitForChange if sw2 isClosed sw2Closed else sw2Opened T start public void swlOpened 42 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE System out println Switch 1 opened T public void sw2Opened 11 System out println Switch 2 opened public void sw1Closed System out println Switch 1 closed if leds 0 isOn 1 System out println Start sampling data leds 0 setColor LEDColor GREEN leds 0 setOn connect2GW try sample catch RecordStoreNotOpenException ex 1 ex printStackTrace catch InvalidRecordIDException ex 1 ex printStackTrace else System out println System is already sampling public void sw2Closed System out println Switch 2 closed if leds 0 isOn 1 System out println Stop sampling data disconnect from GW leds 0 setOff else System out println System is not currently running public void sample throws RecordStoreNotOpenException InvalidRecordIDException StringBuffer sb new StringBuffer String msg empty int i value 1 int count 0 byte outputData null current main loop read 1 sec of data and prints it out System out println Savin
2. is open and listening for a client System out println Radiogram connection is established with Free Range SPOT System out println Waiting for incoming data An Define the date time format to use DateFormat format DateFormat getTimelnstance Variables to write 2 a file FileOutputStream File20utputData null PrintStream MyOutputFile null File20utputData new FileOutputStream HydrophoneValuesFromSPOT csv MyOutputFile new PrintStream File20utputData 48 A 4 HYDROPHONEHOST CODE APPENDIX A SOURCE CODE String to save the data to store on the file String Data2Write String data int count 0 StringBuffer incomingdata new StringBuffer await incoming data while true try Read data received over the radio Utils sleep 50 conn receive dgreply String addr dgreply getAddress read sender s Id System out println Data has been received parse data received data dgreply readUTF incomingdata append data n Data2Write incomingdata toString System out println Received from SPOT count wow count data save hydrophone data to file have we reached EOF if so save to file if lastWord data equalsIgnoreCase EOF MyOutputFile println Data2Write System out println Data has been written in file An reset the stringbuffer incomingda
3. msg length freeMemory A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE T System out println The current battery lvl is System out println The remaining capacity in mwh is count 0 i 0 send remaining data to GW msg sb toString 1 System out println msg to GW sendData msg msg close the recordStore try L rms closeRecordStore catch RecordStoreException ex ex printStackTrace protected void startApp throws MIDletStateChangeException System out println Hello world Battery getBatteryLevel 7 Battery getAvailableCapacity new BootloaderListener start monitor the USB if connected and recognize comma long ourAddr RadioFactory getRadioPolicyManager getIEEEAddress System out println Our radio address JEEEAddress toDottedHex ou System out println The current battery lvl is System out println The remaining capacity in mwh is ADT7411 adc ADT7411 EDemoBoard getInstance getADC enable single channel mode adc write ADCREG byte SINGLECHANNEL AVERAGING OFF analogIn do the heavy stuff before we start sampling to keep up samplerate create or open RecordStore System out println Deleting old RecordStore please wait n try if recordstore exists then delete it so we ha
4. the sensor network doing real time volcano monitoring in Ecuador 9 Also the eProto type board can be used for building our own pre amplifier with complete control over the gain of the amplification and a tight coupling of the hardware parts The benefit of storing the acquired data on the sensor node allows us to finish sam pling before transmitting the data wireless to the GW The result being that we can have a high sampling rate without spending unnecessary clock cycles offloading the data while still sampling When sampling with a high data rate it will take some time to transfer the data We propose a classic duty cycle where the sensor node wake up does its sampling forward the data to the GW do a bit of administrative work like sending information on storage used battery level left etc before going back to sleep The exact length of the sample period will depend on how much battery capacity we can equip the node with while still keeping it inside a buoy 21 3 3 POWER ESTIMATION CHAPTER 3 DESIGN 3 2 2 Designing the CameraSPOT As for building a time lapse sensor node based controller for the camera ourselves several complete guides can be found It will not require any noteworthy computing efforts but rather a pretty accurate time keeping which shouldn t drift too much during the year long deployment Since the arctic circle is on the outskirts of GPS coverage it would be a better solution to sync its time against
5. 1 Pa The origin of these two terms is the intensity of the signal that is transmitted to the water from the target This is called the Source Level SL TL 10log 78 For a plane wave Ip Received signal intensity As the signal travels through the water some of the signal is lost through various propagation losses The totality of this loss is quantified as the Transmission Loss TL As a general rule Transmission Loss is dependent on the distance between the source and the receiver 2 1 TERMINOLOGY CHAPTER 2 HYDROPHONES NL 10log 0 I Noise intensity The Noise Level NL is the sum of the total effect of background and self noise hindering our ability to detect the target signal An average value for the ambient noise level NL of 70 dB is a representative of the shallow water case 4 Figure 2 2 The Passive Sonar Equation 10 Chapter 3 Design In the following we will analyze how to design the hydrophone equipped sensor net work It s important to realize that the hydrophone equipped mote will at some point be part of the existing research installation at Zackenberg and some consideration to the current setup and infrastructure should be taken For instance there is already a buoy in the lake with an Arc rock based IPSerial mote inside 5 which might have space for the hydrophone equipped mote too If not another buoy should be bought for the deployment of the hydrophone into the lake Also there is a
6. 1920s facilitated a detailed depth and ocean bottom 2 1 TERMINOLOGY CHAPTER 2 HYDROPHONES surveys with a speed and accuracy never before available using the traditional lead line techniques to measure water depth Several systems became wide spread and are today used as both scientific instrumentation and indispensable navigation tools 2 1 Terminology Acoustic waves originate from the propagation of a mechanical perturbation Local compressions and dilations are passed from one point to the surrounding points be cause of the mediums the water elastic properties The propagation rate of the pertur bation of the medium is called the velocity But the propagation velocity of an acoustic wave is also imposed by the characteristics of the propagation medium it depends on the density p and the elasticity modulus E 1 0 E C Y p In sea water the velocity of the acoustic wave is close to c 1 500 m s with small variations due to pressure salinity and temperature The density of sea water is ap proximately p 1 030 kg m Fresh water has the density of p 1 000 m so sea water is about 3 more dense than fresh water This compared to air where the respec tive values of sound velocity and density are approximately 340 m s and 1 3 kg m Water has the anomalous property of becoming less dense not more when it is cooled down to its solid form ice It expands to occupy a 9 greater volume in this solid state which acc
7. An Aquarian H2A hydrophone external pre amplifier cables 300 e Harbortronics time lapse camera solution 2 550 e 2 sensor nodes either SunSPOTs 750 or Arch Rock based e A flexible solar panel for the buoy if possible e A hard pipe or similar for housing the hydrophone wire e 1 eFlash and 1 ePrototype add on boards for the SunSPOT e 4 SD cards two for the camera 32 GB and two for the SunSPOTs 2 GB Shttp www wirefreedirect com stand_alone_power _system_design asp 24 3 4 THE HYDROPHONE EQUIPPED SYSTEM CHAPTER 3 DESIGN e Hardware for building our own pre amplifier We estimate that the solution can be built for less than 5000 for the hardware parts The sensor network consists of two sensor nodes one controlling the Aquarian H2A hydrophone and one controlling the shutter control for the time lapse camera The hy drophone node should be placed inside a buoy in the middle of the lake with a pipe protecting the hydrophone wire from the ice The camera node should be placed on the shores of the lake looking out over the lake A solar panel is mounted on the roof of the camera housing prolonging the lifetime of the battery for the camera If possible a flexible solar panel should be installed all around the top of the buoy The hydrophone node is equipped with two stackable add on cards an eFlash card for safe storage of the acquired data on a SD card and an ePrototype board for building our own pre amplifier
8. CHAPTER 3 DESIGN lead acid batteries have a number of drawbacks they are large and heavy and they can vent gases during charge and discharge making them inadvisable to install within a sealed housing Most other secondary rechargeable battery chemistries have high self discharge meaning that they won t work well in a long term application LiPoly batteries however have low self discharge are very light weight and quite compact The Digisnap controller operates on an internal 5V supply but is typically also pow ered by the LiPoly battery pack though the battery converter When the DigiSnap is off there is a load of 8 10 uA on the battery For the lowest power consumption the camera should be configured to not turn the LCD back display automatically on as it is a sure battery drain but also to set the following options e Auto power off 1 minute e Manual focus Shake Reduction Off The solution also includes a 5W solar panel which can be exchanged for a larger panel if necessary At Zackenberg which is located at 74 30 N 21 00 W in North East Greenland 25 km north west of Daneborg there is a partial midnight sun in the period from May 2nd August 12th Its four months of summer and 8 months of winter darkness While the 5W solar panel will be more than enough to recharge the battery during the summer it could be a good design choice to have a larger 20W solar panel for the transition into winter so it s possible t
9. East Greenland Hydrophones are highly versatile sensors with many possible functions as their sens ing capabilities can vary from pure sound monitoring as an aid to the study of marine life and the environment to seismic activity localization as a sonar or as underwater communication devices The design of the hydrophone equipped sensor network will be bound by a number of hard requirements as well as the effect of the arctic deployment environment The solution should be able to operate under the following conditions and constraints e It will be deployed in a freshwater lake in North East Greenland GPS coverage is low in the polar regions e The temperature range will be from 20 to 40 degrees Celcius The depth of the lake is about 6 meters with ice covering the lake during winter with about 2 meters thickness We are looking for a passive hydrophone system We wish to listen after underwater and surface noise not any kind of marine life The solution should be able to operate autonomously for about a year with only one or two maintenance checks a year The solution should provide real time data from the system We need to add a control system to sanity check the acquired sound data in this case a time lapse camera mounted at the shores of the lake And it shouldn t be too expensive 1 2 Contribution The use of sensor networks in aquatic environments has been quite limited partially due to the harsh underwater environ
10. The external pre amplifier from Aquarian mentioned above is for testing the hydrophone before attaching it to the sensor node Battery packs are connected to the two sensor nodes and the camera itself The solution is estimated to be deployed for a whole year with at least one maintenance window scheduled after half a year of operation upon which the SD cards should be swapped and batteries exchanged for new fully charged ones The sensor node controlling the camera wake up once a day to make the camera take a picture Radio contact is established with the GW once a month to sync time The sensor node controlling the hydrophone wakes up every 14 days to make a 45 second sample The data is sent by radio to the GW Data can be transferred by wifi from the GW to the research camp 4 km away Depending on the quality and speed of the satellite link from the research camp data can be forwarded further if possible Y Figure 3 8 Overview of the hydrophone based sensor network setup 25 Chapter 4 Implementation In the following we will take a look at the actual implementation of the prototype for the HydrophoneSPOT and a corresponding GW application the HydrophoneHOST We built the prototype in order to obtain some knowledge on the performance of the system without having access to all the right hardware parts The result being that some of the knowledge gained in this process can be directly transferred to the final system while other par
11. a 9900000000008 NOG 3 5 Figure 3 7 The SunSPOT eFlash and ePrototype add on boards 3 2 1 Designing the HydrophoneSPOT The output from the hydrophone should be connected to one of the sensor nodes Ana log to Digital Conversion ADC input pins The SunSPOT comes equipped with the eDemo sensor board which includes the analog device ADT7411 for ADC conversion The internal oscillator circuit used by the ADC has the capability to output two differ ent clock frequencies This means that the ADC is capable of running at two different speeds when doing a conversion on a measurement channel Setting the ADC into sin gle channel mode in SPOT software will change the default sampling rate from 1 4Hz and make the ADC sample at the specified channel every 44 5 microseconds 22 5 kHz although the SPOT will probably not be able to read it faster than every 142 mi croseconds 7 042 kHz plus whatever loop overhead is needed So the practical max sample rate using the eDemo board will probably be under 6 9 kHz and but it would be better to sample faster than that One option could be to unplug the eDemo board and stack both the eFlash board and the ePrototype board on top of the SunSPOT The eFlash board will be used for storage of the acquired data and the ePrototype board for connecting the analog device TI AD7710 which can sample the input signal at a fre quency of 39 kHz or greater This unit was successfully used in the breakout boards of
12. al al I Yo tan Figure 5 2 Test 2 Running tap Left picture is 20x gain Right picture is with 200x gain The third test shows the signal strength when dropping a small rock into the water next to the hydrophone E tl Lanao roo Figure 5 3 Test 3 Signal strength picture is with 20x gain We also performed the second and third test with the running tap and dropping a rock in water with the hydrophone sensor node and ran the data through Audacity but we were unable to get any recognizable sounds from the sample It might be because the sample rate is too low the signal too weak or it might be due to self noise in the sys tem or all three The most common cause of self noise is ground loops which manifest themselves as a constant 50 Hz tone that doesn t vary with volume In the screen dump from Audacity see below a snippet from a test3 sample can be seen after removal of line breaks from the file and a normalize of the signal 32 CHAPTER 5 TEST AND EVALUATION 000 honeValuesFromSPOT_droprock20 z E L oppg z m i 5 Internal microphone ac OB ata ola 212 22 A E D E Project Rate Hz Selection Start End Length Audio Position 1700 snap To OOHOOmOOs 00h00m16sy 00100 mO2s Click to Zoom In Shift Click to Zoom Out Actual Rate 1700 Figure 5 4 Screendump from Audacity At this point in time it would have been really good to be ab
13. be necessary with the hydrophone as it s sampling at a high data rate so the the SunSPOT sensor node platform from Sun Labs Sun Mi crosystems now Oracle might be worth considering If offers a 180 MHz ARM based CPU with 512 KB RAM and 4MB flash It also offers over the air programming and by the end of the summer 2010 it will have a working IP6 stack on top of the 6LoW PAN protocol It also features different mesh networking protocols The SunSPOT http wwwaarchrock com technology faq php ig http www sunspotworld com http blogs sun com roger entry summer_research_assistants_2010 19 3 2 COMPUTING CHAPTER 3 DESIGN Figure 3 6 The SunSPOT from Sun Microsystems processor board features e 180 MHz 32 bit ARM920T core 512K RAM 4M Flash e 2 4 GHz IEEE 802 15 4 a CC2420 radio with integrated antenna e USB interface e 3 7V rechargeable 720 mAh lithium ion battery e 32 uA deep sleep mode and the general purpose sensor board the eDemo board features e 2G 6G 3 axis accelerometer e Temperature sensor and light sensor e 8 tri color LEDs e 2 momentary switches e 6 analog inputs e 5 general purpose I O pins and 4 high current output pins Also the SunSPOT offers a range of easy to use add on boards allowing us to trans form the SunSPOT for instance into an efficient datalogger by using the eFlash add on board 20 3 2 COMPUTING CHAPTER 3 DESIGN PROTO 2220000000000 REV 1 0
14. false the MIDlet may throw MIDletStateChangeException to indicate it does not want to be destroyed at this time protected void destroyApp boolean unconditional throws MIDletStateChangeHxception 1 for int i 0 i lt 8 i leds i setOff 46 A 4 HYDROPHONEHOST CODE APPENDIX A SOURCE CODE A 4 HydrophoneHOST code Vx SunSpotHostApplication java Sidsel Jensen lt purps diku dk gt HydrophoneHOST app package org sunspotworld import com sun spot peripheral radio RadioFactory import com sun spot peripheral radio IRadioPolicyManager import com sun spot io j2me radiostream import com sun spot io j2me radiogram x import com sun spot util EEEAddress import com sun spot util Utils import java io import java text DateFormat import java util logging Level import java util logging Logger import javax microedition io x ex Sample Sun SPOT host application public class SunSpotHostApplication private static final int RADIOGRAMPORT 37 private static final String SPOTADDR 0014 4F01 0000 1A32 private DatagramConnection dgBConn null private String message empty private RadiogramConnection conn conn2 null private Datagram dgreply null private Datagram data null public void connect2spot try conn RadiogramConnection Connector open radiogram 0014 4F01 D000 1A32 37 dgreply conn
15. normal operating temperature of most digital SLR cameras is specified for OC to 40C Shttp dk farnell com solar panels 17 3 1 SENSING CHAPTER 3 DESIGN The Harbortronics system is an ideal solution for the traditional way of doing remote autonomous monitoring but it does not provide any hooks for real time data retrieval Direct simple connection to s sensor network system is not possible which leaves us with several options Either the Digisnap 2100 can be replaced with a Digisnap 2800 which includes an external input output wire where we can connect the sensor node allowing the sensor node to control the Digisnap which in turn controls the camera but a simpler solution might be to remove the Digisnap controller altogether and let the sensor node and its internal clock control the time lapse sequence of the camera through the shutter release jack To retrieve the data we must also connect a USB cable to the camera A large number of SLR cameras support the Picture Transfer Protocol PTP for bulk transfer which is a widely supported protocol developed by the International Imaging Industry As sociation to allow the transfer of images from digital cameras to computers and other peripheral devices without the need of additional device drivers The protocol has been standardized as ISO 15740 But no sensor nodes that we are aware of support the pro tocol at this point in time PTP is supported on Linux and other free software
16. one ourself and integrate the preamplifier into an add on sensor board placed on the sensor node We will need an add on board anyway since we need to connect the the hydrophone to the sensor node Our recommendation would be to go for a cheaper hydrophone fx the Aquarian H2a It s sturdy cheap and has a low power consumption which makes it ideal for our usage The price for the Aquarian H2a hydrophone does not include pre amplification or cables but it still has a price well below the RESON hydrophone The low price also means that an extra spare can be bought so should the hydrophone be damaged during winter it can be changed at the first possible maintenance check without loosing the possibility of doing measurements the rest of the year 3 1 2 Time lapse camera Our second sensor for the system is a time lapse camera which should be mounted on the shores of the lake Harbortronics sells such a complete time lapse package for Tts also the primary choice for many hobby recording artists which illustrates its ease of use https www harbortronics com Products TimeLapsePackage 14 3 1 SENSING CHAPTER 3 DESIGN Figure 3 3 The Aquarian H2A hydrophone with external pre amp remote monitoring The solution includes Fiberglass Housing glass window High capacity internal Lithium Ion Polymer battery pack 5 Watt Solar Panel Harbortronics Solar Charger Harbortronics Battery Converter Harbortronics D
17. open source operating systems such as FreeBSD by a number of libraries such as libgphoto and libptp as well as the application Gphoto2 In order to get PTP or PTP IP working we ll need a very small scale computer with an operating system and a USB port to transfer the images from the camera to the research station through the wireless connection That could be either a Fit PC2 or Linux on a Gumstix fx the Overo Air model both are probably small and low power enough to fit inside the fiberglas housing But at this point the whole time lapse solution is getting rather complicated with many separate parts and many possibilities for independent failures It might be worth looking for a simpler solution with fewer parts One such simpler solution could be the Campbell Scientific CC640 digital camera for use in harsh environments The CC640 operates at temperatures as low as 40 C and image acquisition can be triggered by applying a5 to 12 volt signal Images are stored on a CompactFlash card or in a dat alogger s memory The images are stored in a JPEG format but unfortunately only in 640 x 480 307 200 pixels resolution which is nowhere near the good resolution of the canon SLR camera The camera is powered by 9 to 15 V with a power drain of 250 mA maximum operating and 250 uA typical quiescent The camera uses the Pakbus pro tocol to transfer images The camera works as a PakBus Leaf node and is not capable http
18. sunspotworld import import import import import import import import import import import import import import import import import import import import import import import XK RA XK XX XX XX E 3 com com com com com com com com com com com com com sun sun sun sun sun sun sun sun sun sun sun sun sun spot spot sensorboard peripheral ISwitch spot spot spot spot spot spot spot spot spot spot spot java io x javax javax javax javax javax javax javax javax javax peripheral Spot sensorboard EDemoBoard sensorboard peripheral ITriColorLED peripheral radio RadioFactory peripheral radio IRadioPolicyManager io j2me radiostream io j2me radiogram peripheral IBattery sensorboard hardware ADT7411 sensorboard io IScalarlnput sensorboard peripheral LEDColor util microedition io x microedition midlet MIDlet microedition midlet MIDletStateChangeException microedition rms InvalidRecordIDException microedition rms RecordStore microedition rms RecordStoreException microedition rms RecordStoreFullException microedition rms RecordStoreNotFoundException microedition rms RecordStoreNotOpenException The startApp method of this class is called by the VM to start the application The manifest specifies this class as MIDlet 1 which means it wi
19. the GW with a fixed interval for instance once a month like its done in the current setup 3 3 Power estimation The lifespan of our application depends on how effective our strategy for using the bat tery will be One of our primary requirements is for the system to work autonomously without human interference for at least 6 months preferably longer As a minimum we need to power the hydrophone the camera and the two controlling sensor nodes including the pre amplifier and the eFlash boards for safe storage of data While the Harbortronics camera solution comes equipped with its own battery and solar panel scaled for long term deployment the hydrophone needs powering either by itself or through the sensor node which calls for an efficient duty cycle The firmware of the SunSPOT enables three different modes of operation Run Basic operation with all processors and radio running Power draw for the eSPOT board is between 70 mA 120 mA The eDemo board can consume up to 400 mA if enabled Idle ARMY clocks and the radio are shut off Idle power consumption is about 24 mA Deep sleep All regulators are shut down except for the standby LDO the power control Atmega and pSRAM Deep sleep power consumption is 32 yA Table 3 2 Overview of the SunsPOTs three modes of operation Deep sleep cannot be entered if if the radio is on if external power is supplied or if USB power is on Deep sleep and idle modes can b
20. the surface Snowfall on the other hand has a muffling effect and the sound can only travel to a limited extent The ice sheet acts as a huge membrane across which the cracking and popping sounds spread 2 1 3 The Passive Sonar Equation The sonar equations are founded on the basic equality between the desired signal and undesired background portions of the received signal For a sonar to successfully detect an acoustic signal we require that it is above a detectable threshold DT Signal Level gt BackgroundLevel Signal to noise ratio is an important concept because it represents the degree to which an amplifier can be successfully employed to improve the situation If the signal to noise ratio S N or SNR is too low the noise is nearly equal to the signal In this case amplification will also increase the noise and provide no substantial improvement For high signal to noise ratios amplification will improve the magnitude of the signal relative to the noise We can express this correlation with the passive sonar equation SNR SL TL NL DI where SL is the source level TL is the transmission loss NL is the noise level and DI is the directivity index The directivity index DI for our network is zero because we assume omnidirectional hydrophones All the quantities in the equation are in dB re uPa where the reference value of 1 Pa amounts to 0 67 x 1072W atts cm SL 10log s Is Signal Intensity Ip Reference Intensity
21. water ice interface are not well known 3 the measurement of under ice contours is difficult and 4 the diffraction of sound around ice obstacles may be important Due to scattering at the rough boundaries of the ice only low frequencies typically less than 40 Hz can propagate to long distances in the Arctic channel 7 2 1 TERMINOLOGY CHAPTER 2 HYDROPHONES 2 1 2 Causes of noise Noise is an important component of underwater acoustics The cause of noise can be grouped into four categories e Ambient noise typically originates from factors outside the system and stems from natural or man made causes e Self noise typically from the systems own electronics e Reverberation typically caused by parasite echoes but only affects active sonar systems e Acoustic interference typically generated by other acoustic systems nearby The two first categories are the ones most likely to cause interference in our system By definition ambient noise is the noise received by the sonar in the absence of any signal and self noise from the system Ambient noise has several very distinct physical origins each corresponding to particular frequency ranges e Seismic and volcanic activity very low level frequencies e Shipping and industrial activity from 10Hz 1kHz e Surface agitation depends on sea state and wind speed from a few hertz to a few tens of kHz We don t anticipate any noteworthy seismic and volcanic activity
22. 8 0 002 duty cycle finding a bigger battery or trying to harvest some renewable energy But unfortunately it doesn t end there we must also consider the transmission time of the samples if we sample at 7kHz for 60 seconds and each data point in the sample is one byte we get 7000 x 60 x 1 420 kBytes per minute According to the CC2420 radio datasheet the radio has a transmit speed of 250 kbps 420 kBytes per minute x 8 bit 250 kbit per second 13 44 seconds which means we have to reduce the sample period from 60 seconds to about 45 sec onds if we are to preserve energy for both sampling and transmitting the data But preferably we would like to sample with 44kHz with the Aquarian H2A hydrophone which will give us even more data to transfer 44000 x 60 x 1 2 640 kBytes per minute 2 640 kBytes per minute x 8 bit 250 kbit per second 84 48 seconds so for each one minute sample it will take us 1 minute and 24 seconds to transfer the data That would require that we reduced the sampling period by more than half If we still want a 1 minute sample it would have to be once every fourteen days or so We suggest that we possibly try to find a bigger 3 7V Li ion battery for the SunSPOTs 23 3 4 THE HYDROPHONE EQUIPPED SYSTEM CHAPTER 3 DESIGN 3 3 1 Energy harvesting In order to prolong the lifetime of the application we need to consider using the pos sibility for energy harvesting Both solar panels and wind turbi
23. GW installation on the shore of the lake with a working wireless connection to the research station some 4 km away A side from that the proposed solution will be autonomous from the existing one to allow maximum independence in case of system failures 3 1 Sensing We wish to design and build a sensor network system centered around the input from two sensors a hydrophone deployed in the lake and a time lapse camera mounted on the shore of the lake The sensor node controlling the hydrophone will wake up at timed intervals and make sound recordings which will be sent to the GW In order to be able to control and match the sound recordings against some ground truth we wish to take at least one time lapse picture a day on which the current weather con dition around the lake can be seen The time lapse camera provides our eyes in the field during the remote monitoring mission It is our hope that we later during the post processing of the sound recordings can match the recordings to a certain weather situation and a specific noise in the recording It might also be a good idea to mount a small weather sensor so we can pickup on for instance wind speed or very large differences in temperature So if the wind speed stays above a certain threshold for a longer period which could be a favorable condi tion for our sound monitoring we wish to wake up our sensor nodes to perform extra 11 3 1 SENSING CHAPTER 3 DESIGN measurements We canno
24. Hydrophone equipped mote by Sidsel Jensen Department of Computer Science University of Copenhagen June 2010 Abstract Underwater sensor networks is a novel area within sensor networks Recently there has been a growing interest in underwater wireless sensor networks due to its advan tages and benefits in a wide spectrum of applications in aquatic environments The main challenges of deploying such a network are the traditional ones of cost and lim ited battery resources of individual sensor nodes but also consideration to the water environment and the difficult access possibilities once deployed As a result both hard ware and software must be carefully designed In this project we present the design and implementation of a small scale prototype for a hydrophone equipped sensor network for long term deployment in the arctic area Contents 1 Introduction 1 1 Application constraints A an AA AAA TR Contribttion 2 2 7 e ea ae Sivas Sods Bowe amp eS eee A VS CNIS A bos ts SE Bp tec ice FE A eon bed Bt re eh Relate 2 Hydrophones 21 Terminology x a tne C T RR E hs awn Sarl Oe eaters A Selene ee eal ZA Propagation LOSSES s e cR T Bt ew di ai Dil Dv ACAUSES OF NOISE A RR ec A A a 2 1 3 The Passive Sonar Equation a Se oe oh NS tue oS 3 Design Oe Sensing o e Mat BU Be de oc A te hl Sone at e 3 1 1 Choosing a hydrophone viii gee eee A ee 31 2 Dane lapsecametas 40 one ole Oy cel OM ey ee ea 32 COMPU tat vg ea e
25. ailures or misconfigurations occur its not possible to detect it before the re trieval of the instruments e The amount of data is limited by the capacity of the onboard storage devices Underwater sensor nodes and networks enable real time applications for oceano graphic data collection pollution monitoring and offshore exploration but the water also introduces a number of challenges for sensor network systems such as low com munication bandwidth large propagation delays floating node mobility and high er ror probability The lifetime of the application is limited by the battery capacity and the limited energy harvesting possibilities solar energy and wave generation And the underwater sensors themselves are prone to fouling and corrosion 1 1 Application constraints To make matters worse the environmental conditions in the arctic area are harsh Sensor networks primarily has four types of sensor activities sensing transmitting receiving and computing which are all put to the test in extreme weather conditions MANA http www ece gatech edu research labs bwn UWASN http www itu dk mana 1 2 CONTRIBUTION CHAPTER 1 INTRODUCTION is a research project with the goal of improving scientific data acquisition in polar re gions The goal of this project is to explore the possibility of adding a hydrophone equipped sensor network mote to the current MANA research installation located in a freshwater lake in North
26. ate the distance between the sonar and its target and receiving the signal on a suitable antenna completes the measurement with a determination of the angle of arrival of the signal Active sonars are also called projectors Passive sonars are designed to intercept noises and possible active sonar signals radiated by a target vessel Their main interest lies in their total stealth Passive sonars are also called hydrophones The first underwater echo detection systems were developed by British American and French scientists for the purpose of underwater navigation by submarines in World war I and in particular after the Titanic sank in 1912 Their two primary functions was to locate submarines and icebergs but these earliest systems only worked at relatively high frequencies and achieved detection ranges of only several thousand yards un der favorable conditions Between the two world wars the sonar technology improved considerably It benefitted from the emergence of first generation electronics and from progress in the newborn radio industry In the aftermath of World War II the Cold War between the Western and the Eastern blocks ensured continued focus on the efforts in the scientific and technological research on underwater acoustics 2 In parallel with military developments oceanography and industry were able to profit from the development of underwater acoustics For instance the invention of the sonic depth finder SDF in the early
27. ble to build the designed system despite the problems we have experienced with our prototype but it will require a serious effort to obtain the needed duty cycle with the limited battery capacity to meet the deployment criteria Also it will require a detailed plan for several longer field tests before the actual deployment Underwater wireless sensor networks does provide a number of advantages and ben efits for aquatic applications but the design and implementation calls for meticulous ground work and demands the utmost of the sensor nodes limited and constrained hardware 34 Bibliography 1 An introduction to underwater acoustics principles and applications Xavier Lur ton Springer 2003 2 The Past Present and Future of Underwater Acoustic Signal Processing Jos M F Moura Carnegie Mellon University IEEE Signal Processing Magazine July 1998 3 Sensor Networking in Aquatic Environments Experiences and New Challenges ThiemoVoigt et al Second IEEE International Workshop on Practical Issues in Building Sensor Network Applications 15 18 Oct 2007 Dublin Ireland 4 Battery Lifetime estimation and optimization for underwater sensor networks Raja Jurdak Cristina Videira Lopes and Pierre Baldi University of California Wiley Interscience 5 Monitoring in a High Arctic Environment Some lessons from MANA Marcus Chang and Philippe Bonnet 2009 6 Autonomous Hydrophones at NOAA OSU and a New Seafloor Sentry Syst
28. c 100k Figure 4 3 The LM386 based pre amplifier 4 1 3 Connecting the hardware Three wires are connected from the breadboard to the SunSPOT It receives power from the 5V pinout on the SunSPOT and it s also connected to the ground pinout The amplified output from the hydrophone is connected to pinout AO on the SunSPOT 00000000 Figure 4 4 The connection to the SunSPOT http wwwosepino com circuits index mini_amplifier_lm386 jpc 28 4 2 PROTOTYPE LIMITATIONS CHAPTER 4 IMPLEMENTATION Once the program is deployed to the SunSPOT sampling can be initiated by press ing the leftmost button The first LED in the row will light green to show that sampling has begun The program can be stopped by pressing the rightmost button When send ing data to the host program the second LED flashes red for each package sent The use of the LEDs makes it easy to identify the operation of the SunSPOT 4 2 Prototype limitations Unfortunately we have not had access to either an eFlash or an ePrototype add on board during development We have been restricted to use the on board 512K RAM and 4MB flash of the SunSPOT In order not to run out of memory on the SunSPOT it was nescesarry to overwrite the data once it was sent to the host program We had to change the main sample loop to gather data and send it as soon as we had enough data to fill a datagram instead of finalizing all sampling before sending it to the host pro
29. close to our sys tem and shipping will not pose a problem as the hydrophone is to be deployed in a freshwater lake Surface agitation on the other hand will prove very interesting to measure Air bubbles are created by sea surface movements for instance by heavy rain and cause ambient noise They form an inhomogeneous layer close to the surface whose dual phase mixing modifies locally and strongly the acoustic characteristics of the propaga tion medium velocity and attenuation This process decreases in importance as depth increases because the hydrostatic pressure increases Below 10 20 meters the effect of bubbles can be neglected In the mid frequency band 200 Hz 50 kHz the dominant noise source is wind acting on the sea surface It has been shown that there is a strong correlation between ambi ent noise and wind force or sea state Ambient noise increases about 5dB as the wind strength doubles Peak wind noise occurs around 500 Hz and then decreases about 6dB per octave 2 1 TERMINOLOGY CHAPTER 2 HYDROPHONES Frozen lakes are known to give off most noise during major fluctuations in tempera ture the ice expands or contracts and the resulting tension in the ice causes cracks to appear Due to the changes in temperature the hours of morning and evening are usually the best times to hear these sounds Thin ice is especially interesting for acous tic phenomena it is more elastic and sounds are propagated better across
30. d up the headphones and desoldered the existing electronics inside the headphones and then soldered the electret microphones onto the same wires We then tested the microphones with the normal input jack on a computer using a mic input Line in will not work as electret microphones require a little power to function We then filled the headphones with silicone and pulled the microphones back into the head phones and wiped the excess silicone off and left it to dry Afterwards we cut the mini jack off the wire in order to use the wires directly on the breadboard Figure 4 2 The DIY hydrophone 4 1 2 Building the pre amplification We started out with a very simple transistor based pre amplifier from Nerdkits but unfortunately it was very unstable and it only enhanced the signal 6 7 times which http leafcutterjohn com p 915 http www freesound org forum viewtopic php p 13253 7 http www nerdkits com videos sound_meter 27 4 1 BUILDING THE PROTOTYPE CHAPTER 4 IMPLEMENTATION wasn t enough So instead we chose a pre amplifier layout based on the LM386 unit The LM386 is a power amplifier designed for use in low voltage consumer applications The gain is internally set to 20 to keep external part count low but the addition of an external resistor and capacitor between pins 1 and 8 will increase the gain to any value from 20 to 200 The amplified output can be read from pin 5 x200 4 184 X20 4 18 10k Mi
31. e entered through the programming Waking the processor up from deep sleep can be done with either an alarm an external interrupt or by pressing the attention button Let s do a rough power estimate The runtime power draw from the hydrophone equipped sensor node will worst case be around 120 mA 400 mA 520 mA but much can be done to reduce this number If the goal is a duty cycle of 90 10 90 sleep 10 awake spread across a year http www glacialwanderer com hobbyrobotics p 13 22 3 3 POWER ESTIMATION CHAPTER 3 DESIGN 10 of 365 days 36 5 days 36 5 days x 24 hours 876 hours 365 days 2 4 hours per day 2 4 hours x 520 mA 1248 mAH The capacity of the built in Lithium ion battery is 720 milliampere hours so we would deplete the battery by using the SunSPOT for 1 38 hours Now let s try to do the calculation the other way around Lets assume we can have the SunSPOT turned on for a total of 1 hour over a period of half a year after which the battery can be changed during maintenance 60 minutes x 60 seconds 180 days 20 seconds per day If we only sample once every five days we could do a 1 minute sample So our sample period will be 60 seconds 1 hour x 520 mA 520 mAH for the whole period That leaves 200 mAH on the battery for deep sleep mode 180 days x 24 hours 4320 hours 4320 x 32 A 138 mAH which is within the 200 mAH limit That leaves us with the task of programming either a 99 99
32. em for Real time Detection of Acoustic Events H Matsumoto et al IEEE OCEANS 2006 7 Challenges Building Scalable Mobile Underwater Wireless Sensor Networks for Aquatic Applications Jun Hong Cui ACM 2007 8 An Underwater Network Testbed Design Implementation and Measurement Zheng Peng et al ACM 2007 9 Fidelity and Yield in a Volcano Monitoring Sensor Network Georff Werner Allen et al 10 SunSPOT Developers Guide Sun Microsystems Sun Labs August 2008 11 SunSPOT Owners Manual Sun Microsystems Sun Labs August 2008 12 SunSPOT Theory of Operation Sun Microsystems Sun Labs August 2008 13 Arch Rock IPSerial Node Product Datasheet 14 User Manual for Aquarian H2A hydrophone 35 BIBLIOGRAPHY BIBLIOGRAPHY 15 Datasheet for RESON TC4032 hydrophone 16 Datasheet for Bruel and Kj r 8103 hydrophone 17 User manual for Digisnap 2100 18 ChipCon CC2420 SmartRF datasheet 36 Appendix A Source code A 1 Arduino code The Arduino code define ANALOGIN 0 void setup Serial begin 9600 void loop int val analogRead ANALOGIN Serial print Oxff BYTE Serial print val gt gt 8 Oxff BYTE Serial print val Oxff BYTE 37 A 2 PROCESSING CODE APPENDIX A SOURCE CODE A 2 Processing code x Oscilloscope Gives a visual rendering of analog pin 0 in realtime This project is part of Accrochages See http accroc
33. g values to RecordStore n while true for i 1 i lt 7000 i try add value to RecordStore value analogIn getValue 43 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE Utils sleep 5000 msg Integer toString value try values i rms addRecord msg getBytes 0 msg getBytes catch RecordStoreException ex ex printStackTrace values i analogIn getValue count System out println count mow value pack data in bundles of 15 samples and send try outputData rms getRecord i catch RecordStoreException ex ex printStackTrace T if i 6999 1 String data new String outputData sb append data EOF else String data2 new String outputData sb append data2 if i gt 0 amp i 15 0 flash led 2 while sending data leds 1 setColor LEDColor RED leds 1 setOn msg sb toString System out println msg to GW sendData msg sb delete 0 sb length leds 1 setOff wow i L catch IOException ex ex printStackTrace try make delimiter between samples rms addRecord finish getBytes 0 finish getBytes length catch RecordStoreException ex ex printStackTrace status on the spot System out println Free Memory left on SPOT 44 Runtime getRuntime
34. gram As a consequence our sampling rate is lower than it should be We start over writing old data after 7000 samples To make matters worse the SunSPOT have had a problem with enabling the single channel mode Once enabled the read values don t change at all The problem might be that we are trying to read the ADC faster than it is sampling but as long as this bug persists we are unable to sample faster than 1 7 kHz public static final byte ADCREG byte 0x19 public static final byte SINGLECHANNEL byte 0x10 public static final byte AVERAGING_OFF byte 0x20 IScalarInput analogIn EDemoBoard getInstance getScalarInputs EDemoBoard A0 ADT7411 adc ADT7411 EDemoBoard getInstance getADC enable single channel mode adc write ADCREG byte SINGLECHANNEL AVERAGING OFF analogIn getIndex toInteger 4 3 Data post processing The HydrophoneHOST program simply waits and listens for data Once received it saves the received hydrophone data to a file on disk The file can be imported into Audacity as a RAW file Audacity is a free easy to use and multilingual audio editor and recorder that works on a range of operating systems Files can be exported to a wealth of formats e Encoding Unsigned 8 bit PCM http audacity sourceforge net 29 4 3 DATA POST PROCESSING CHAPTER 4 IMPLEMENTATION e Byte order No endianess e Channels 1 channel MONO e Start Offse
35. hages drone ws x c 2008 Sofian Audry info sofianaudry com x Modified by Sidsel Jensen to save data to file import processing serial x PrintWriter output Serial port Create object from Serial class int val Data received from the serial port int values void setup T size 640 480 Open the port that the board is connected to and use the same speed 9600 port new Serial this Serial list 1 9600 values new int width smooth output createWriter output_microphone txt int getY int val 1 T return int val 1023 0f height 1 void draw while port available gt 3 if port read Oxff val port read lt lt 8 port read output println val An T for int i 0 i lt width 1 i values i values i 1 values width 1 val background 0 stroke 255 for int x 1 x lt width x line width x height 1 getY values x 1 width 1 x height 1 getY values x 38 bps A 2 PROCESSING CODE APPENDIX A SOURCE CODE void keyPressed if key s output flush Write the remaining data output close Finish the file exit Stop the program T J 39 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE A 3 HydrophoneSPOT code RX XX StartApplication java Sidsel Jensen lt purpsQdiku dk gt HydrophoneSPOT app package org
36. igiSnap 2100 Canon Rebel XS 1000D camera A pair of 8 GB memory cards All required tools cables manuals and accessories The enclosed camera is a Canon Rebel XS 1000D which shoots with 10 1 Megapixel with a maximum resolution of 3888x2592 pixels There s the choice of two lower resolu tions and all three sizes can be recorded with either Fine or Normal JPEG compression Its also possible to record RAW files either with or without a Large Fine JPEG Best quality Large Fine JPEGS typically measure between 3 5 and 4 MB each while RAW files measure around 10MB each Shooting in RAW format and saving them on the 8 GB SD card gives us room for about 800 pictures but if we choose Large Fine JPEGs we have room for more than twice as many pictures roughly 2000 With the RAW format we can do two pictures a day with the Large Fine JPEGs we can take five pictures a day for a whole year without filling up the card entirely But SD cards comes cheap these days so we would recommend buying a pair of 32 GB cards leaving plenty of room 15 3 1 SENSING CHAPTER 3 DESIGN Figure 3 4 The Harbortronics Time Lapse package for pictures Energy tests in the lab should be performed to see how much power is consumed by powering on the camera taking a picture and shutting it down again to measure what is the safe limit for how many pictures can be taken if the solution is to operate at least 6 months without human interference The perhaps mos
37. ishery echo sounders and sonars Acoustic Doppler current profilers lann Sediment profilers Seismics Figure 2 1 Frequency ranges of the main underwater acoustic systems 1 intensity of any kind of flux through a medium for attenuation Because the propagation medium is limited by the sea surface and the seabed the signals transmitted undergo reflections at the interfaces Any given signal can therefore propagate from source to receiver along several distinct paths multiple paths The main signal arrives along with a series of echos of amplitudes decreasing with the number of reflections undergone It is called backscattering when the reflection of an acoustic wave collide with an obstacle in the water or at the limits of the medium and reflects the acoustic wave and send an echo of the signal back to the direction they came from It is a diffuse reflection as opposed to a specular reflection like a in a mirror The time structure of the signal is somewhat affected and the performance of a system can be highly degraded by these parasite signals Depending on the application the echoes can be either desirable or undesirable if they are completely jamming the useful signal In general there are four factors that are peculiar to the arctic environment which com plicate the modeling of acoustic propagation 1 the ice keels present a rapidly varying surface 2 the reflection transmission and scattering properties at the
38. ke free from the ice cover which is roughly 2 meters thick during the winter The lake is about 6 meters deep so this would put the hydrophone right in the middle between the water surface and the lake bottom The two expensive hydrophones are built to withstand deep un derwater pressure of 500 m depth or more Since we are doing our measurements in what may be characterized as shallow water the extra endurance of the expensive hy drophones might not be necesarry However the hydrophone might still get trapped in the ice during spring when the ice melts and refreezes and extra care needs to be taken to protect the wire connecting the hydrophone to the mote as the ice can be very sharp The wire should be packed in an isolating pipe which shouldn t make too much noise when crushed to relief some of the pressure from the ice which will form around it during the winter A good design choice could be to go for one of the cheaper type of http www afabsound com home php http www dolphinear com de specs htm Shttp www reson com sw3154 asp http www bksv dk Products TransducersConditioning AcousticTransducers Hydrophones 12 3 1 SENSING CHAPTER 3 DESIGN hydrophones or build one ourselves since there is a high risk of loosing it during winter Figure 3 1 The DolphinEAR De100 hydrophone Looking at the coating of the four hydrophones all three but one looks very sturdy The DolphinEAR is not packed in the
39. le to substitute the self made hydrophone for a professional one in order to eliminate some of these problems We are quite sure the hydrophone works as seen when using the Arduino but the Arduino transfers the data through USB which is significantly faster than the current sample rate of the SunSPOT With a lower sample rate we lose a lot of data With a professional hydrophone at hand we could at least have aligned the samples to find a baseline Once the professional hydrophone is available it could also be very interesting to do some serious power consumption tests for the system These tests would prove the basis for a recalculation of the duty cycle If we in any way could reduce the power consumption in run time and combine it with a bigger battery pack we might have the possibility to wake up more often to sample 33 Chapter 6 Conclusion In this project we have explored the possibility of building and deploying a hydrophone equipped sensor network mote and we have designed and implemented a working pro totype of a SunSPOT based mote Unfortunately we ran out of time while debugging the results of the tests conducted with the self made hydrophone A minimum require ment is a faster analog to digital conversion chip if the SunSPOTs are to keep up with the sample rate of the hydrophone Also a better and stronger professional hydrophone is needed in order to establish the actual performance of the system We believe it is feasi
40. ll be selected for execution public class StartApplication extends MIDlet public public public public public static final byte ADC REG byte 0x19 static final byte SINGLE CHANNEL byte 0x10 static final byte AVERAGINGOFF byte 0x20 static static final int SAMPLEPERIOD 30 how many seconds final int SAMPLE RATE 7000 at 7k rate 40 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE public static final int NO SAMPLES SAMPLE PERIOD SAMPLE_RATE private String message empty private String BSAddr 0014 4F01 0000 10C2 private DatalnputStream in null private DataOutputStream out null private DatagramConnection dgBS null private boolean talkedtoBS false RecordStore rms int recordStoreIndex 0 to measure battery level IBattery Battery Spot getInstance getPowerController getBattery to use the Leds private ITriColorLED leds EDemoBoard getInstance getLEDs ISwitch swl ISwitch sw2 EDemoBoard getInstance getSwitches EDemoBoard SW1 EDemoBoard getInstance getSwitches EDemoBoard SW2 IScalarInput analogIn EDemoBoard getInstance getScalarInputs EDemoBoard A0 can use store about one second of data private int values new int 7000 public void connect2GW try while talkedtoBS false System out println No contact to GW establish connection if dgBS n
41. ments and associated high system costs Hy drophones as a sensor have only recently been integrated into underwater sensor net work installations and testbeds despite the fact that they are well known and powerful tools for leveraging the task of localized sound monitoring underwater We present the design and implementation of a small scale prototype for a hydrophone equipped sensor network for long term deployment under the extreme weather condi tions in the arctic 1 3 OUTLINE CHAPTER 1 INTRODUCTION 1 3 Outline In the first part of this project we review a small but relevant part of the terminology regarding hydrophones Chapter 2 and describe and analyze the design space and cor responding design choices Chapter 3 In the second part we discuss the implemented prototype Chapter 4 and conducted tests Chapter 5 and in the third and final part we will evaluate and conclude on the findings Chapter 6 Chapter 2 Hydrophones A transducer is an electronic device that converts energy from one form to another whereas a SONAR SOund NAvigation and Ranging is a technique which uses sound under water to navigate or detect other vessels The sonar works after the same echo principles as a RADAR RAdio Detection And Ranging Sonars are classified into two main categories depending on their mode of functioning Active sonars transmitting a signal and receiving echoes from a target The mea sured time delay is used to estim
42. nes have been installed with success on bases in Antarctica proving their sturdiness in a harsh off grid environ ment The most obvious choices would be e solar energy e wave power generation or e wind power Unfortunately wave power generation is not currently a widely employed com mercial technology and the lake is covered by ice all winter making it impossible to effectively use this strategy for most of the year Also setting up a wave generator in the lake in close proximity to our buoy could generate a lot of noise disturbing our samples Setting up a wind turbine near the lake is also a possibility but the force of the arctic storms might take it out A solar panel will work great during the summer months but produce little or no power during the dark winter even though the snow on ground will work as a reflector for the daylight The solar panel is probably the best value for money solution and in fact there already is a solar panel in the current in stallation powering the GW during summer but a hybrid approach combining a wind turbine with a solar panel would probably give a better all year yield However the solution needs to be within close proximity to our sensor node inside the buoy which renders using a wind turbine practically impossible but a small flexible solar panel might be mounted on the sides of the buoy 3 4 The hydrophone equipped system The proposed sensor network system consists of the following parts e
43. newDatagram conn getMaximumLength System out println Awaiting connection attempt from SPOT conn receive dgreply message dgreply readUTF System out println Received message if message equalsIgnoreCase SPOT signing in Utils sleep 5000 47 A 4 HYDROPHONEHOST CODE APPENDIX A SOURCE CODE data conn newDatagram conn getMaximumLength data writeUTF ACK conn send data System out println Sent ACK to client message empty dgreply null else System out println Something went wrong i catch IOException e 1 System out println No connection finally 1 conn close public String lastWord String s String lastWord int spaceIndex s indexOf if spacelndex 1 lastWord s else lastWord s substring s lastIndexOf 1 System out println Last word in message is lastWord return lastWord xx x Print out our radio address public void run throws FileNotFoundException Exception 1 long ourAddr RadioFactory getRadioPolicyManager getIEEEAddress System out println Our radio address IEEEAddress toDottedHex ourAddr initialization of connection variables for radiogram communication wait for SPOT to contact us connect2spot System out println Port RADIOGRAMPORT
44. ng devices that vary in complexity and capability and they prevent the excessive discharge of the battery which would damage the battery and protects electronic appliances against over voltage They can turn off any device operated through it when the battery voltage drops below a critical point On the other 5 http www elfrasolen dk Produkter Batterivagt index htm 13 3 1 SENSING CHAPTER 3 DESIGN hand the battery might be the cheapest part of the whole design and we want to keep on sampling for as long as absolutely possible even if that means depleting the battery completely It might be a better solution to choose a smaller dynamic frequency range and a small sample rate to reduce the power drain Rubber front panel snaps into grooves in battery cover with double seal Figure 3 2 The Aquarian UP1 pre amplifier inside The signal we receive from the hydrophone will most likely need some sort of pre amplification otherwise it will be too weak Usually a professional external amplifier is used to enhance the signal further but including that in the on site setup is not recom mendable It would require even more space for hardware and even more power The RESON hydrophone comes with a built in 10dB pre amplifier and with the Dolphin EAR a small external amplifier is included but mostly the hydrophone suppliers lets you buy the pre amplifiers as an extra accessory Instead of buying one we could also build
45. o recharge the batteries even with fewer hours of light The solar charger that comes included in the time lapse package can be reused Harbortronics recommends that the solar panel rests on top of the housing where it can also serve as a rain shield to minimize drops on the front window of the housing With a larger solar panel that might not be possible if the solar panel should be placed in the optimal 45 degree angle Apart from shielding the wires from wildlife the user guide gives no other indications as how to shield the box from the arctic environment A snow storm might possibly cover the window with snow and ice leaving the camera blind It will most likely be worthless to isolate the box itself as the cold would still penetrate the glass window but isolating the battery could prove a good investment for the battery lifetime The Harbortronics solution seems very stable The university of Alaska has tested the system at low temperatures in their facilities and they found that the system worked all the way down to the lowest tested temperature 60C However at 40C and below some pictures were missing from the sequence Some of the pictures were dark others half light half dark suggesting that the mechanical items in the camera the shutter and the mirror assemblies may have been sticking at times or otherwise slowed The timing never varied suggesting that all of the electronics worked at all temperatures even though the
46. ounts for the fact of ice floating on liquid water Acoustic signals are characterized by the number of vibrations per second the fre quency f expressed in Hertz The frequencies used in underwater acoustics range roughly from 10 Hz to 1 MHz depending on the application For a sound velocity of 1 500 m s the underwater acoustic wavelengths will be 150 m at 10 Hz 1 5 mat 1kHz and 0 0015 m at 1 MHz 2 1 1 Propagation losses The propagation of a sound wave is associated with an acoustic energy expressed as either the acoustic intensity I or the acoustic power P Intensity and power can vary enor mously A high power sonar transmitter may deliver acoustic power of several tens of kilowatt whereas a nuclear submarine in silent mode might radiate only a few milli watts When acoustic waves propagate the most visible process is their loss of intensity because of geometric spreading a divergence effect and absorption of acoustic energy by the propagation medium itself This propagation loss or transmission loss is a key parameter for acoustic systems Sea water is a dissipative propagation medium it absorbs part of the energy of the transmitted wave We also call the gradual loss in 2 1 TERMINOLOGY CHAPTER 2 HYDROPHONES Frequency kHz 0 1 10 100 1 000 Maximum ranges km 1 000 100 10 l 0 1 Multibeam sounders Sidescan sonars Transmission and positioning Active military sonars Passive military sonars F
47. same hard cover as the rest of the hydrophones its piezoelectric membrane is much more exposed which means it might be unfit for an arctic deployment The missing coating will also make it much more exposed to corrosion during the year long deployment compared to the other three When listening for noise it s a good idea to choose the broadest possible frequency bandwidth but that costs power The DolphinEAR is the hydrophone with the small est frequency span whereas the Briiel and Kj r hydrophone has the broadest frequency span They are all however within the frequency span which is relevant for noise ob servations They all claim to have low self noise which is important when listening for ambient noise We want to be able to hear the signal it shouldn t drown in electro static noise from the system itself You normally apply a high pass filter over the recording to remove the noise but this might not be possible as we then might also remove the valid noise signal we are listening for The power requirement is highly relevant for the lifetime of the application Increas ing the intensity and the pre amplification of the hydrophone signal requires higher power The higher power drain combined with the arctic environment with below 0 temperatures for half a year at a time will most likely deplete the battery faster than we anticipated A device called a battery guard might be worth considering into the setup These are voltage monitori
48. t O bytes e Amount to import 100 e Sample rate 1600 Hz We have not had the time to look into different signal processing schemes for re duction of unwanted noise in the received signal Further data post processing will most likely be necessary before any valid information can be obtained from the data samples 30 Chapter 5 Test and evaluation We have performed initial testing of the hydrophone and the sensor node but longer tests in a more realistic setting are required Both regarding the deploy depth but also regarding the performance in sub zero temperatures The initial testing with the Ar duino showed that the self made hydrophone was waterproof and responded in water Preliminary tests conducted in a kitchen sink showed that the 200 times gain on the pre amplifier introduced too much noise in the system With the 20 times gain we get a clearer but fainter signal but most importantly it isn t drowned in noise Test 1 shows it very clearly The signal to noise ratio is too low It would have been good if we had the possibility to slowly increase the gain on the pre amplifier to find the right level of pre amplification E A A SEES nay o loa U Figure 5 1 Test 1 Noise level in semi quiet environment Left picture is 20x gain Right picture is with 200x gain 31 CHAPTER 5 TEST AND EVALUATION Kitchen sink test with a running tap right above the hydrophone which was placed in a bucket underneath i T
49. t afford energy wise to do this very often but it will most likely provide us with recordings of extreme situations which can help set the normal recordings into the right context and give us a larger data set which will provide more detail and precise information Preferably the data set from the first 6 months should be used to debug and fine tune the system for the next deployment period 3 1 1 Choosing a hydrophone For the sake of simplicity we have selected 4 hydrophones for consideration two in the cheap range the Aquarian H2A hydrophone and the DolphinEAR and two in the expensive range a RESON TC4032 and a Bruel and Kj r 8103 Miniature Hy drophone They are all omni directional and passive listening devices All four are from well known companies and very reliable Aquarian DolphinEAR RESON Briiel and Kj r Description H2A DE100 Series TC4032 8103 Miniature Frequency 10 Hz 100KHz 7 Hz 22 kHz 5Hz to120kHz 0 1Hz to 180 kHz Power 0 3 mA Approx 7 mA at 9V 19mA at 12VDC 6mA without load Sensitivity 180dB re 1V uPa 170dB re 1V pPa 211 dBre1V pPa Connector 3 5 minijack 3 5 minijack LEMO 10 32UNF microdot plug Pre amplifierlinternal imp buffer amp external PA amplifier low noise 10dB external amp Cable not included 8m 6m 10m Price 159 319 DKK 17 887 00 Table 3 1 Comparison of 4 different hydrophones We will be doing measurements about 3 meters down in the la
50. t important device apart from the camera itself is the Digisnap 2100 which controls the camera trigger The DigiSnap can be configured for either A Simple Time Lapse STL sequence which consists of an initial delay followed by a number of pictures taken with a particular interval between each picture or an Advanced Time Lapse ATL sequence where a set of Time Lapse sequences can be programmed to start at particular times of the day but the DigiSnap controller does not have a real time internal clock so once powered up it will presume that it s midnight In terms of comnectivity a flap on the left side of the camera body opens to reveal the TV output a socket for the optional RS 60E3 remote switch and a USB port The Digisnap controller connects to the shutter release jack the RS 60E3 pinout The interface is very simple generally just a switch contact The Digisnap controller is configured by connecting a null modem cable between the controller and a PC Mac and using a simple terminal program Energywise the package comes equipped with a single high capacity Lithium Ion Poly mer LiPoly rechargeable battery pack having a nominal voltage of 11 1V and 9AH ca pacity The fully charged LiPoly battery pack has enough capacity for about 2 months of operation between charges depending on the details of the application The most com mon battery chemistry for long term remote applications is lead acid Unfortunately 16 3 1 SENSING
51. ta delete 0 incomingdata length catch Exception e System err println Caught throw e e T xx Start up the host application param args any command line arguments public static void main String args throws Exception try SunSpotHostApplication app new SunSpotHostApplication app run catch FileNotFoundException ex Logger getLogger SunSpotHostApplication class getName log Level 49 while reading sensor samples SEVERE null ex A 4 HYDROPHONEHOST CODE APPENDIX A SOURCE CODE 50
52. ta eR AE Boe AD A an A ke 3 2 1 Designing the HydrophoneSPOT oe ey eee 4 aS 3 2 2 Designing the CameraSPOT 8 AAA oh eee 33 Power estimation Ps Ba be oe A ee eas 33 1 Energy h rvestng negak ae kep ey eos 3 4 The hydrophone equipped system aaau 4 Implementation 4 1 Building the prototype ie A iaa 4 1 1 Building the hydrophone roce a 4 1 2 Building the pre amplification o oo 4 1 3 Connecting the hardware ooo o 4 2 Prototype limitations ade Sok Be ia 4 3 Data post processing rra a Bo ay a EA eae A Boe 5 Test and evaluation 6 Conclusion ii PWN HA N ODD UI 11 12 14 19 21 22 22 24 24 26 26 27 27 28 29 29 31 34 CONTENTS CONTENTS A Source code 37 AGL ZATAUINO Code a Ai ta Boh ik e far tog 37 AZ PLOCESS SCOOTER AE ta Bee eae amp 38 A 3 HydrophoneSPOT code o E amp NE A A Ge amp 40 AA HydrophoneHOST code 2 itn ae ete BESS AA 47 Chapter 1 Introduction The surface of our globe is covered by 70 percent water yet doing underwater data acquisition is cumbersome The traditional approach for remote ocean monitoring is to deploy underwater sensors that record data during the monitoring mission and then at a later point recover the instruments but this approach has a number of disadvan tages e Real time monitoring is not possible e No interaction is possible between onshore control systems and the monitoring instruments e If f
53. ts will have to be redone once the final system is complete 4 1 Building the prototype We built the prototype using the Arduino Duemilanove board and a breadboard The Arduino is an open source electronics prototyping platform based on flexible easy to use hardware and software and using it allowed us to quickly build the hardware while still being able to test it The finished hardware on the breadboard was then transferred from the Arduino to the SunSPOT where the actual code for the application was written Figure 4 1 The Arduino duemilanove prototype board http arduino cc en Main ArduinoBoardDuemilanove http www arduino cc 26 4 1 BUILDING THE PROTOTYPE CHAPTER 4 IMPLEMENTATION 4 1 1 Building the hydrophone A lot of people build their own hydrophone using a simple piezo electric material a pre amplifier and a housing but using a normal electret microphone in this case a Panasonic WM 61A works just as well The Panasonic WM 61A electret microphone support 20 20 kHz frequencies An electret microphone is one of the best value for money omnidirectional microphones you can buy Electret microphones can be very sensitive very durable extremely compact in size and has low power requirements Our cheap do it yourself DIY hydrophone consists of the following parts e a cheap pair of headphones e 2 Panasonic WM 61A electret microphones bought on eBay and e silicone for making it watertight We opene
54. ull dgBS DatagramConnection Connector open radiogram BSAddr 7 37 Datagram dg dgBS newDatagram dgBS getMaximumLength sendData SPOT signing in System out println Sent SPOT signing in to basestation Utils sleep 5000 dgBS receive dg hang here wait for ACK from BS message dg readUTF System out println Received message while message equalsIgnoreCase ACK salenge den kun modtager tomme beskeder sa heng her Utils sleep 1000 dgBS receive dg hang here wait for ACK from BS message dg readUTF System out println While loop received message if message equalslgnoreCase ACK System out println Received message from GW 41 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE talkedtoBS true else System out println Wrong initialisation msg from basestation catch IOException e System out println No route to 0014 4F01 0000 10C2 finally 1 T synchronized public void sendData String m try Datagram dgreply dgBS newDatagram dgBS getMaximumLength dgreply writeUTF m dgBS send dgreply catch IOException ex ex printStackTrace public void startSW1WatchThread new Thread public void run while true swl waitForChange if sw1 isClosed sw1Closed 1 else sw1Opened
55. ve room for new try if RecordStore listRecordStores null RecordStore deleteRecordStore SampleData System out println Done T catch RecordStoreException ex 1 ex printStackTrace System out println Creating new RecordStore please wait n rms RecordStore openRecordStore SampleData true System out println Done catch RecordStoreException ex rAddr Battery getBatteryLevel Battery getAvailableCapacity getIndex toIntege samples ex printStackTrace T 45 A 3 HYDROPHONESPOT CODE APPENDIX A SOURCE CODE start program by pressing swl startSW1WatchThread stop program by pressing sw2 startSW2WatchThread System out println Press left button to start sampling sync time with host after last data sent send battery and time status go into deep sleep duty cycle T protected void pauseApp 1 This is not currently called by the Squawk VM xx Called if the MIDlet is terminated by the system I e if startApp throws any exception other than MIDletStateChangeException if the isolate running the MIDlet is killed with Isolate exit or if VM stopVM is called It is not called if MIDlet notifyDestroyed was called param unconditional If true when this method is called the MIDlet must cleanup and release all resources If
56. www gphoto org http www fit pc com http www gumstix com http www campbellsci com cc640 digital camera specifications Bhttp www campbellsci com documents manuals pakbusnetguide pdf 18 3 2 COMPUTING CHAPTER 3 DESIGN Figure 3 5 The Campbell Scientific CC640 digital camera of performing any routing The camera could be connected to a sensor node through the cameras RS 232 output but in order to retrieve the images the sensor node would have to understand the Pakbus communication protocol to fetch the images Which solution to go for depends on whether the captured images should be retrieved in real time or not If we just want to control when the pictures are taken but not re trieve the pictures until the first maintenance check the Harbortronics solution would be perfect However if the image data needs to be retrieved periodically we would recommend to look for a different solution 3 2 Computing The existing sensor network at Zackenberg uses the Arc Rock IPserial nodes for com puting The benefit of the Arc Rock nodes is that they offer a whole range of ICMP control and management messages as well as support for full UDP and TCP transport layer protocols as well as over the air OTA programming For the easiest integration it would be best to choose a similar node for the new setup However the chip driving the IPserial node is not particularly strong and some pre processing on the node might

Hydrophone equipped mote

Contents

Download Pdf Manuals

Related Search

Related Contents