An Introduction to Wireless Networking for Ecological Research John Porter, University of Virginia & Thomas Williams, AirNetworking.com.

An Introduction to Wireless Networking for Ecological Research

John Porter, University of Virginia

&

Thomas Williams, AirNetworking.com

John Porter – Jan. 2004

Why Network?

Ecological research has been conducted for many decades, often in sites remote from even basic telecommunications. Why do we need to employ networks at ecological research sites?


Why Network?

Network access to field sites is desirableTo reduce logistical costs associated with collecting data and putting it into electronic formsImprove data quality by providing real-time feedbackTo provide access to Internet resources (e.g., online keys, maps) while in the field


Why Network?Network access to field sites is critical to:

“Observing the unobservable” – allowing us to answer questions that are otherwise unapproachable

Providing temporally intensive and spatially distributed dataNeeded to capture rare eventsNeeded to deal with distributed processesHeterogeneous sensor arrays


Why Network?

Networked “robots” make feasible complex experimental manipulations that remain under the control of the investigator


Why Wireless?

Ecological research sites are typically inaccessible or prohibitively expensive for wired connections

Because we can! (now)


Wireless 101

Basic approaches to wireless networking include:Low-speed telephony-based modems Cell phones (costly if you need many) Satellite Phones (costly)

Moderate-speed wireless serial connections May include a central unit that polls peripheral

units on a fixed schedule Up to 115 KB/s

High-speed wireless ethernet (2 to 100 MB/s)


Wireless Technology

Traditional radio modems have concentrated in sending a single strong signal resulting in:High power consumptionRelatively low transmission speedsRequire FCC licenses to avoid interference


Spread SpectrumStarting in 1986, spread spectrum was declassified It was co-invented in the 1941 by actress Hedy Lamarr

Unlike traditional radios, it uses a broader range of radio frequencies, but at very low powerSpread spectrum uses only a subset of the band at a particular time, allowing multiple signals (even from multiple sources) to coexist simultaneously without significant interference.


Spread Spectrum

Spread spectrum radios use relatively low powerCuts power requirementsReduces risk of interference – also no FCC

license is required


Spread Spectrum Frequencies

The FCC allows unlicensed spread-spectrum radios on 3 major bands900 MHz2.4 GHz5 GHz

All are strictly “line of sight” with maximum ranges of up to ~30 kmVegetation and metal tend to block signals


900 MHz

Not as widely usedMay not be available in all countries (e.g.,

not Europe, not most of Africa or Latin America)

Does better than other bands at penetrating vegetation

Used in the popular “Freewave” serial spread spectrum radios


2.4 GHzMost commonly used band for 802.11b,g “Wi-Fi” wireless ethernet in office environments

Poor penetration of foliageUses frequencies similar to those used in a

microwave oven – water converts radio energy to heat!


5 GHz

Used for 802.11a wireless ethernet (increasingly popular)

Intermediate ability to penetrate foliage

Equipment typically somewhat more expensive than 2.4 GHz


Examples of Use

North Temperate Lakes LTER Buoy Network – Serial Spread Spectrum

Trout Lake Station

Buoy~3 km

Raft

~2 km

Graphic by Tim Kratz and Paul Hanson


Photos: Paul Hanson & Tim Kratz

Base Station

900 MHz radio


Uses

Real-time publication of data on the web

Remote control of Buoy functions

Allows addition of Sonar & Imagers

Apprise Technology


Examples of Use

Virginia Coast Reserve LTERBarrier Island system Islands are isolated from conventional

(wired) telecommunications


The VCR/LTER uses a hybrid network with both proprietary 900 MHz and standard WiFi 802.11b 2.4 GHz wireless Ethernet connections.

Areas within line of sight of two towers are tinted in yellow

900 MHz2 Mb/s

802.11b11 Mb/s

= VCR/LTER Lab

VCR/LTERWirelessBackbone


Uses of Wireless at VCR/LTER

Real-time Meteorological & Tide data (3 networked stations currently deployed)Web Cameras (6 currently deployed)Access to networked data resources (e.g., the web) in the field

Integrated camera/ web server/radio/power


Uses of WebcamsCapture time series

Education Non-obtrusive observationObserve rare events

“A picture is worth a thousand words”


The Art of Wireless

Wireless to Wired EthernetBridge (WET11)

Internet

Field

Webcam

WirelessAccess Point

Lab

Basic System for 802.11b


The Art of Wireless


Internet

Field

Wired Hub

WebcamSerial-to-EthernetConverter

Data Logger orA/D converter


Lab

Expanded System for 802.11b

Add data logger


The Art of Wireless


Internet

Field

Wired Hub

WebcamSerial-to-EthernetConverter

Data Logger orA/D converter


Lab

Long-Range System for 802.11bAmplifier

Directional Antenna

Added: amplifier and directional antenna


Approximate Ranges

Omni Directional

Omni 100 m 500 m

Directional 500 m 1 km

No Amplification Used


Approximate Ranges

Omni - amplified

Directional

amplified

Omni 1 km 10 km

Directional 10km 20 km

Amplification on one end


Approximate Ranges

Omni - amplified

Directional

amplified

Omni- amplified

~10-20 km 20 km

Directional amplified

20 km ~30-50 km

Amplification on both ends


CostsWireless to Ethernet Bridge $100

Wireless Access Point $60

Directional Antenna $75

Freewave Radio $1200

2.4 GHz Amplifier $330

Serial to Ethernet Converter $150

Combo Wireless to Ethernet Bridge and serial converter

$300

Pigtail Cable adaptor $20

Solar Panel (55 watt) $300


The Challenges

Obtaining a line-of-sight unobstructed by vegetation, ground or metal buildings is keyEven if direct line-of-sight is possible, the

Fresnel effect may prevent communications if part of the signal is blocked, so towers are desirable


Possible Solutions

Use towers

Use wired connections to instrumentation in the vicinity of a tower

Use frequencies that better penetrate vegetation (e.g., 900 MHz)

Relay signals around obstructions (“mesh” network)


Power System

Solar Panel

Solar Controller

Battery

120 v Inverter

Digital Timer

Power Strip

Radio Ethernet Hub

Cameras

Sensors

Data Logger


Sample WebcamWWW cameraWireless Bridge

12-120 v Power Inverter

Digital Timer

Power adapters

Power strip

Ant

enna


Challenge: Power

Providing power (especially 24/7) can be difficult

Equipment often require different voltages


Possible SolutionsConcentrate power-hungry equipment (e.g., amplifiers) at the laboratory and deploy only low power equipment in the fieldUse efficient DC-AC inverters or DC-DC converters to deal with different voltagesUse timers or remotely-controlled data loggers to run the system only when needed


Some lessons learned

Power supplies, not radios, are the most difficult componentMost consumer-grade DC-DC voltage

converters are power hogsUse cheap inverters, not expensive ones

The cheap ones reset automatically if batteries are drawn down, expensive ones don’t….

Use digital, not analog timers to cut down on hours of operation to save power

Cheap inverters have poor frequency control


Beyond?!!

Here we have focused on using commercial, off-the-shelf (COTS) technologies that are relatively inexpensive and available todayAt the smaller scale, substantial research is being dedicated to the development of small, autonomous “motes” that can be used to create self-configuring networks of sensorsThere are also higher-powered, licensed microwave systems that can cover longer distances and carry higher data rates


Questions?

An Introduction to Wireless Networking for Ecological Research John Porter, University of Virginia & Thomas Williams, AirNetworking.com.

Documents

network access

wireless networking

comwhy network

fieldwhy network

investigatorwhy wireless

ecological research

low powerspread spectrum

spectrum radios2