Tutorial about Seismic Sensor Network

Vinayak Naik, Martin Lukac, and Deborah Estrin

Information Processing in Sensor Networks (IPSN’07), Cambridge, MA

April 24, 2007

Acknowledgments to Igor Stubailo, Derek Skolnik, Joey Degges, and Mike Allen for lending us equipments and time.

Special demands of seismic and acoustic applications

• Seismic– Large-scale deployment spanning hundreds of kilometers– It’s not easy

• Highly varying links with frequent disconnections results in challenged networks

• Remote monitoring and fixing of nodes demands services such as reliable broadcast, sink-based data collection, and maintenance of a global state

• Developing these services become non-trivial due to challenged networks

• Acoustic localization– Sampling rate of the order few KHz– Lew will summarize the challenges

Outline

• Using the seismic array out-of-the-box

• A few words about seismology

• Remotely managing and configuring array

after the deployment

• Assembling the array in 30 minutes

• Adapting the software to fit your needs

What’s in the box?

• 1 PC• 3 Cens Data Communication Controller

(CDCCs) • 1 Q330 (a combined ADC and data logger)• Ubuntu live CD, which contains

– Emstar source code– Emstar code compiled for the i366 and stargate

architectures– TFTP server and minicom to reflash the nodes (to

be used while assembling the array)– You may also use the CD to install all the required

software on your PC or run it in an emulator such as qemu!

Using the CD

• Prerequisites:– A computer that can be booted using a CD and has wired

ethernet connection

– A basic knowledge of Linux, such as use of ssh, scp, and ifconfig

• Procedure:– Boot your computer using the CD

– Set password for ubuntu: "sudo passwd ubuntu”

– setup IP address for the ubutu: “ifconfig eth0 131.179.145.X

netmask 255.255.255.0 broadcast 131.179.145.255”

– If using a virtual machine, unload USB-to-serial driver if alread

loaded

The seismic activities before the start of the tutorial

• Stop the data collection process (Duiker)

• Transfer data to the base station (PC)

• Strip the DTS header from the packet

• Uncompress the data

• Convert the data from miniseed to ascii format

• Transfer data to your laptop

• Plot the data using gnuplot

Wait, the theory about seismology is coming up.

• Start Duiker and let it run for 4 minutes

• Stop Duiker

• Strip the header

• Uncompress the data

• Convert the data from miniseed to

ascii format

• Transfer data to my laptop

• Plot the data using gnuplot

In situ data collection and presentation

Same as the previous slide

Outline

• Using the seismic array out-of-the-box

• A few words about seismology

• Remotely managing and configuring array

after the deployment

• Assembling the array in 30 minutes

• Adapting the software to fit your needs

Seismology 101

P-wave speed: 1.5-8 Km/sS-wave speed: 60-70% of the speed of P-wave

Wikipedia: An earthquake is a phenomenon that results from the sudden release of stored energy in the Earth’s crust that creates seismic waves.

There are two types of seismic wave, 'body wave' and 'surface wave'. There are two kinds of body waves: primary (P-waves), travel fastest through any type of matterand secondary (S-waves), shear, the most destructive.

Body waves travel through the Earth’s interior:

Seismic wave energy

Richter TNT for Seismic Example Magnitude Energy Yield (approximate) -1.5 6 ounces Breaking a rock on a lab table 1.0 30 pounds Large Blast at a Construction Site 2.0 1 ton Large Quarry or Mine Blast 4.0 1,000 tons Small Nuclear Weapon 4.5 5,100 tons Average Tornado (total energy) 6.5 5 million tons Northridge, CA Quake, 1994 7.0 32 million tons Japan Quake,1995;Largest Thermonuclear Bomb 8.0 1 billion tons San Francisco, CA Quake, 1906 9.0 32 billion tons Chilean Quake, 1960 12.0 160 trillion tons Fault Earth in half through center

160 trillion tons of dynamite is a frightening yield of energy. Consider, however, that the Earth receives that amount in sunlight every day.

Because of this huge amount of energy released the seismic waves travel large distances and make possible to capture them with different kinds of seismic sensors (seismometers).

Seismic sensors

Most signals are composites of many frequencies.

Analog with light and sound:Seismic Light SoundShort-period Blue TrebleLong-period Red Bass

Typical seismogram

The long-period and short period instruments are called "narrow" band used for volcano experiment by Harvard. They sense frequencies near 1/15 s and 1 hertz respectively. The yellow region is the low end of the frequency range audible to most humans, 20 hertz to 20,000 hertz.

A broadband instrument senses most frequencies equally well.For our data collection we use the best in its class CMG-3T broadband sensor, made by Guralp Systems.Its standard frequency response is 120 s – 50 Hz what results in high quality seismic data.

Frequency responsesof seismometers

About Middle America Subduction Experiment (MASE)

• We have a seismic deployment to study the structure of the mantle in Mexico

• The deployment consists of wireless stations covering large distances

• We developed software to:– Handle collection the seismic data– Manage the seismic system

• This tutorial presents this software and how to use it

Seismic deployment application requirements

• Extensive: 500 Km from Acapulco through Mexico City to Tampico

• Dense: 1 sensor every 5-10 Km• High bandwidth: Data acquisition rate: 3 - 24 bit

channels at 100Hz each• Online and Reliable: Semi real-time (on the order

of days), reliable data delivery to UCLA for analysis

• Online system management– Query state, change configuration, update binaries– Can not interfere with data delivery

• Application driven topology: application determines sensor placement

– Infrastructure does not (Can’t rely on pre-existing cell or power infrastructure)

50 standalone Caltech sites62 wirelessly connected UCLA sites

MASE: Given these requirements, we deployed solar powered seismic stations equipped with 802.11b

MASE wireless seismic station

15 dBi YAGI or 24 dBi Parabolic 2.4GHz antenna

70 watt solar panel, GPS

mast and guy wires

Quanterra Q330 24-bit digitizer

sensor controller

2.4GHz amp

car battery

CDCC (CENS Data Communication Controller)

Guralp 3T seismometer

A block diagram of the system’s architecture

DTS & file_mover

TCP/IP, UDP

Duiker

Q330 (ADC)

Sensor

WiFi

CDCC

Replace with your own

ethernet

Pakistan earthquake

Our network: •Achieves almost 10 times better resolution than the previous network as of Oct. 2005 (with 50 sites total). Now it is 20 times better (100 sites)

•Provides visualization of the upper mantle and the subduction process, coast to coast across Mexico.

Google video

• The data was used to analyze the structure of

the earth underneath Mexico

• Results are being submitted to the Science

journal

Outline

• Using the seismic array out-of-the-box

• A few words about seismology

• Remotely managing and configuring array

after the deployment

• Assembling the array in 30 minutes

• Adapting the software to fit your needs

Networking support needed for both data acquisition and system management

• Data delivery – Bandwidth driven– Bandwidth: 20-40 of MB per day per station– Latency: get the data eventually, but reliably– Many to one routing

• System Management – Latency driven– Bandwidth: usually less than 10’s of KB’s– Latency: as fast as possible– One to all routing and back

Use of wireless network for remote operation

• Demonstrate use of Delay Tolerant Shell (DTS)– Start dtsh– Issue a ps command– See result of the ps command

• Demonstrate the use file transfer– Xfer a file from /opt/test

• Demonstrate the use of file mover– Create a file on a stargate– Show the same file on the PC

• Xfers– Shows the active transfers

• Links– Shows existing links on a node

• Sink_status– Shows the upstream route to the sink

Configuration utilities

Data collection utility

Adjunct utilities

Challenges handled by DTS, file transfer, and file mover

• Frequent unpredictable disconnections– Rainy season: sites flood (some

24x7), trees grow– Wind: misaligned antennas– Equipment malfunction: amps

burn, voltage regulators break

• Poor and unstable links– Connectivity secondary concern

for site selection– Stretched links highly

susceptible to weather and environment

• Useful tools for operating wireless sensor networks under harsh wireless settings

B

System management

• Existing management tool: remote shell (ssh)

• Modified management tool: Disruption Tolerant Shell– Asynchronous remote shell to all

nodes in network simultaneously

– Provides node management capabilities when end-to-end connections are unavailable or fail

– Ensures that commands will succeed: as long as there is eventually a connection between a node and any other node that already has the command

A

DC

E

F

df –hls /opt/dts/file_mover | wc

Commands

Responses

Data delivery using DTN techniques

• Buffer data into hour long bundles (1-3 MB)• Deliberate one hop bundle transfer• Path to sink determined by best ETX• Improvement over end-to-end

– Not affected by path disconnections– Keeps retrying on single link instead of full path– Continual ‘progress’ being made towards sink– More efficient use of bandwidth in face of

disconnections and bottlenecks

A

B

C

F

end-to-end

hop-by-hop

Extra fun features of DTS

• Guaranteed in order execution from source node

• Reboot and crash safe• Implicit feed back on nodes and

links: spot bottlenecks, dead nodes• Execute a command on individual

nodes• Push a file to all nodes

– Distribute new script or component

Handling sessions in DTS

• A sequence number is assigned per source node per session• Each node publishes a ‘starting sequence number’ across the

network– It identifies the oldest command issued by a node that should be in

the network– Any commands and responses with sequence numbers below the

value (for that particular node) are discarded and not propagated• User controls the starting sequence number

– To remove commands from the network, user increments the commands source node starting sequence number

– Can choose to do this after all the nodes have reported responses or sooner

• Giving control of seqno to user is simple, easy to understand, and efficient

• Utilities to handle seqno– Use seqno command to see all the nodes starting sequence

numbers– Use incr command to increment the starting sequence number on

the current node

Outline

• Using the seismic array out-of-the-box

• A few words about seismology

• Remotely managing and configuring array

after the deployment

• Assembling the array in 30 minutes

• Adapting the software to fit your needs

Ingredients

• 3 stargates to form a 1-hop network

• 1 computer

• 1 serial cable

• 1 ethernet hub and 1 ethernet cable

Assembling a seismic node

• Connect an episensor to the Q330

• Connect Q330 to the wired ethernet hub

• Connect a stargate to the wired ethernet

• Connect wireless antenna to the stargate

• Note that you can substitute Q330 with your

choice of data logger

Reprogramming the stargates

• Connect PC to the wired ethernet

• Connect a serial cable from PC to a stargate

• Configure minicom profile called “stargate0”

• In stargate-install.exp, change the IP address of the

TFTP server to PC’s IP address

• Flash the kernel and the root file system– The kernel and the root file system comes with all the

seismic software!

– Screenshot of the flashing in progress

Configuring a gateway node (base station)

• Designate a stargate as a gateway

• Restart DTS

Index

• Episensor– Measures movement across multiple axes

• Q330– Data logger, GPS, accurate maintenance of time

• PDA– Reports status and configures Q330 via infra-red

• Williard– A closed-source software to retrieve the data from Q330

• Duiker– An open source software to retrieve the data from Q330– A comparison with Antelope (supports network, open source,

and inexpensive)

• DTS– An open source software for the remote management of

stargates

Outline

• Using the seismic array out-of-the-box

• A few words about seismology

• Remotely managing and configuring array

after the deployment

• Assembling the array in 30 minutes

• Adapting the software to fit your needs

Use of the software for other wireless sensor networks

• Replace Q330 with ADC of your choice

• Install a driver that collects data from the ADC

and creates files on the stargate at

/opt/dts/xfer

• file_mover will transfer files to the gateway

node

• No change in DTS and other utilities

Convert existing 7.2/7.3 stargates into seismic nodes

• Download dts-whole-system.tar.gz and

dts-whole-system-install.tar.gz to /opt on the

stargate

• Make sure that the script dts-whole-system-

install.tar.gz is executable

• Execute the script

Adapting the DTS code for your needs

• Change code in

emstar/devel/dts/dts/dts_status.c

• Compile code for stargate architecture

• Stop DTS if it is running

• Copy the new code to the right place on a

stargate

• Start DTS and see the change

Convert other platforms into seismic arrays

• Portable to Linux-based platforms

• Instructions to port EmStar to other platforms

Seismology of the future at CENS

• Deploy the CDCCs in Peru• Use of low power LEAP-II nodes instead of

stargate• Use of low power and inexpensive ADC

boards from Reftek Corp. instead of Q330• Deploy combination of the LEAP-II and the

new ADC – For GeoNet to study aftershocks– For structural health monitoring of tall buildings in

Los Angeles

A few upcoming features of DTS

• Provide visualization of the data movement

– Using a coarse grained global time (one second), recreate ‘movie’ of file movement for entire network

– Can help spot network problems and bottlenecks

• Upload data to SensorBase.org– Makes it easy to visualize and browse

data collection status– RSS feed can provide access to anyone

who wants to monitor problems or generic status of network

• Web interface to simplify operation– Command line interface is nice for Linux

pros– Web interface more intuitive for

asynchronous model

Thank you

Wish you happy seismography!

• Resources for users and developers– Emstar web-page– Emstar mailing list– Disruption Tolerant Shell in the Proceedings of the

2006 SIGCOMM workshop on Challenged Networks

Use of seismic sensing

• The similarity between the Mexico and LA region

• P and S waves• How is the seismic array different from the

Harvard's volcano motes?• What is the sampling frequency

Need for DTS, file transfer, and file mover

• Unreliable links• Need to broadcast commands to the nodes

and get responses from the all the nodes• Need to broadcast files to the nodes• Hop-by-hop data movement

%18 - A%152 - B%69 - C%77 - D%107 - E%42 - F%81 - G%202 - H%76 - I%106 - J%95 - K%53 - L%157 - M

D

M

NG

E

C

F

B

L

A

K

H

I

J

A – sinkDirect inetconnection

13 Node Cuernavaca Line

• Network topology does not reflect the mostly linear physical topology

Data paths