Wilamowski B.M., Irwin J.D. The Industrial Electronics Handbook. Second Edition: Industrial Communication Systems

Подождите немного. Документ загружается.

Ultralow-Power Wireless Communication 10-5

hidden-terminal.problem,.some.protocols.also.neglect.it.and.use.direct.channel.access,.assuming.that.col-

lisions

.are.unlikely.due.to.rare.wake-ups.[EO08].. e.resulting.network.protocols.are.quite.simple.from.

a.WSN.perspective,.but.very.energy.ecient.for.the.sensor.nodes,.as.long.as.the.always-on.coordinator.

is.not.taken.into.account.

.the.coverage.area.is.large,.then.multi-hop.approaches.are.more.energy.ecient.than.single-hop.

approaches.[NPK06],.because.the.signal.strength.falls.in.a.logarithmic.manner.making.short.communi-

cation

.distances.signicantly.more.energy.ecient.. ese.multi-hop.approaches.require.specic.rout-

ing

.approaches.and.coordinated.wake-ups.of.the.nodes.forwarding.messages..Simple.CSMA.protocols.

are.not.sucient.for.that.and.more.sophisticated.MAC.protocols.need.to.be.used,.which.are.discussed.

later.on.in.this.section.

Full

.exibility.of.the.network.is.usually.reached.with.mesh.networks.. ey.add.a.topology.control.

protocol.layer.allowing.easy.integration.of.mobile.nodes..Other.protocols.provide.a.self-healing.capabil-

ity

.of.the.network.oen.taking.advantage.of.redundant.nodes..Some.approaches.also.consider.enhanc-

ing

.the.network.lifetime.through.redundant.nodes.that.mainly.sleep.for.the.rst.part.of.their.life.to.take.

over.the.job.of.nodes.that.have.used.up.their.energy.resources.[TG02].. ese.redundancy.approaches.

improve.network.reliability,.but.require.higher.hardware.costs..Nonetheless,.all.these.protocols.add.an.

overhead.of.management.messages.to.the.communication.that.require.energy.for.transmission.and.keep.

nodes.awake.

.low.duty-cycle.is.most.important.for.ultralow-power.devices.. e.communication.needs.to.be.coor-

dinated

.in.a.way.that.still.allows.most.of.the.network.to.be.asleep..Several.issues.should.to.be.avoided.

during.communication.like.idle listening.to.the.channel.if.no.other.node.intents.to.send.a.message.or.

overhearing.messages.not.addressed.to.the.node..Also,.nodes.should.avoid.transmitting.messages.to.

sleeping.receivers.(overemitting)..Two.or.more.senders.should.avoid.sending.messages.at.the.same.time.

on.the.wireless.channel.that.causes.collisions.corrupting.messages..As.a.result,.an.ideal.communication.

protocol.ensures.that.sender.and.receiver.are.only.awake.for.message.transmission.and.sleep.the.rest.of.

the.time.. is.is.manageable.as.long.as.only.two.nodes.are.involved.and.both.work.on.a.periodic.sched-

ule

.with.synchronized.clocks..In.multi-hop.networks,.where.a.node.may.need.to.forward.messages.of.

other.nodes,.this.requires.suitable.protocols.that.are.discussed.in.the.following.paragraphs.

Depending

.on.how.the.wake-up.intervals.of.nodes.are.coordinated.in.the.network,.the.protocols.

can.be.distinguished.in.on-demand,.asynchronous,.and.rendezvous-based.approaches.(Figure.10.5)..

On-demand.approaches.oen.use.a.second.low-power,.low-rate.radio.to.wake-up.a.node.for.communi-

cation

.[STS02].. is.approach.is.limited.in.practice,.due.to.the.cost.for.the.second.radio.and.its.range.

dierence.to.the.main.radio.as.less.energy.is.used.

.asynchronous.approaches,.the.sleep.cycles.of.nodes.are.not.synchronized.. e.common.approach.

is.that.a.sender.waits.until.a.receiver.is.awake.to.transmit.the.message.. erefore,.nodes.periodically.

wake-up.and.check.for.messages..B-MAC.is.the.most.commonly.used.protocol.in.which.a.sender.starts.

with.a.long.preamble.[PHC04]..Other.nodes.waking.up.on.their.periodical.schedule.recognize.this.

(b) (c) ReceiverReceiverData channel

SenderSender

Check

Data message

Wake-up channel

Wake-up message

Check

Synchronized

wake-upPreamble

(a)

FIGURE 10.5 Illustration. of. typical. wake-up. approaches. for. MAC.. (a). On-demand,. (b). asynchronous,.

and(c)rendezvous.based.

10-6 Industrial Communication Systems

preamble.and.stay.awake.to.receive.this.message.. e.drawbacks.of.this.approach.are.overemitting.for.

the.prea

mble

.and.over

hearing

.at.node

.not.invo

lved

.(add

ressed)

.in.the.comm

unication

.as.they.need.to.

stay.awak

.unti

.the.prea

mble

.ends.to.chec

.the.mess

age.

.More.ener

.ecie

.are.X-MA

.[BYA

H03]

and.CSMA

-MPS

.[MB0

4],

.wher

.mult

iple

.shor

.mess

ages,

.with.the.rele

vant

.node.addr

ess,

.are.repe

ated

inst

ead

.of.a.sing

.long.prea

mble.

. is.not.only.avoi

.the.over

hearing

.of.node

.that.are.not.invo

lved,

.but.

also.allo

.inte

rested

.node

.to.indi

cate

.when.they.are.awak

.and.read

.to.rece

ive.

. e.down

side

.of.thes

asyn

chronous

.appr

oaches

.is.that.the.ener

.probl

.is.main

.shi

.to.the

sender,

.who.need

.to.send.

the.prea

mble.

. e.leng

.of.the.prea

mble

.depe

nds

.on.the.chec

.inte

rval

.of.the.rece

iver

.and.the.ener

ecie

ncy

.beco

mes

.a.trad

e-o

.betw

een

.over

emitting

.at.the.send

.and.idle.list

ening

.at.the.rece

iver.

Rendezvous-based

. appr

oaches

. sync

hronize

. the. wake

-up

. peri

ods

. of. a. grou

. of. node

. e. rst.

appr

oach

.was.the.S-MA

.proto

col

.[YHW

04],

.in.whic

.all.node

.awak

.peri

odically

.for.a.xed.time.to.

exch

ange

.mess

ages.

.IEEE.802.

15.4

.uses.a.comp

arable

.appr

oach

.in.its.beac

on-enabled

.mode

.As.in.both.

appr

oaches,

.node

.not.invo

lved

.in.comm

unication

.also.stay.awak

.for.the.acti

.time

;

.this.resu

lts

.in.

over

hearing

.and.idle

-listening

.agai

.T-MA

.[DL0

.impr

oves

.this.issu

.by.sett

ing

.unin

volved

.node

to.slee

. e.down

side

.of.all.rend

ezvous-based

.appr

oaches

.is.that.all.node

.wake

-up

.at.leas

.for.the.

sync

hronization,

.and.the.ener

.cons

umption

.depe

nds

.stro

ngly

.on.the.wake

-up

.cycl

.whic

.is.usu-

lly

.cons

tant.

.Espe

cially

.in.scen

arios

.with.a.dyna

mic

.comm

unication

.load

.this.lead

.to.idle.list

ening,

when.no.node.has.a.mess

age

.to.tran

smit,

.and.coll

isions,

.when.mult

iple

.node

.inte

.to.send.mess

ages.

 e.adapt

ion

.of.the.wake

-up

.cycl

.to.the.tra

.load.of.the.node

.is.ther

efore

.one.appr

oach

.to.redu

energy-consumption.while.providing.a.high.throughput.[NPK05].

Rendezvous-based

.appr

oaches

.are.usua

lly

.more.ener

.eci

ent

.than.asyn

chronous

.appr

oaches.

.But,.

appl

ication

.requ

irements

.shou

.also.be.consi

dered

.when.deci

ding

.abou

.comm

unication

.proto

cols.

For.exam

ple,

.if.a.temp

erature

.sens

.samp

les

.a.new.valu

.ever

.minu

te,

.then.the.node.shou

.not.have.

to.sync

hronize

.in.a.rend

ezvous-based

.MAC.with.a.node.samp

ling

.the.illu

mination

.ever

.100.ms..In.

this.case

.an.asyn

chronous

.appr

oach

.may.be.a.bett

.choi

ce,

.but.then.the.prea

mble

.for.the.temp

erature

sens

.migh

.be.very.long

.Henc

.netw

ork

.topol

ogy,

.comm

unication

.proto

cols,

.and.appl

ication

.desig

need.to.be.har

monized

.for.ult

ralow-power

.com

munication.

10.4 application Layer approaches

An.appropriate.application.design.is.essential.for.ultralow-power.communication..Again,.the.application.

requ

irements

.of.the.node.need.to.be.cons

idered

.to.choo

.the.most.ener

gy-aware

.desi

gn.

.Two.gene

ral

desi

.prin

ciples

.shou

.be.used

.rst

.allo

.the.node.to.slee

.as.oen.and.as.long.as.poss

ible;

.and.seco

nd,

proc

ess

.info

rmation

.loca

lly,

.as.the.data.proc

essing

.consu

mes

.sign

icantly

.less.powe

.than.tran

smitting

data.[RSS

02].

. us,.the.numb

er of transmitted messages,

.the.mess

age size

.and.the.numb

er of wake-ups

shou

.be.redu

ced,

.whil

.stil

.prov

iding

.the.requ

ired

.func

tional

.qual

ity.

.For.exam

ple,

.if.the.netw

ork

.is.

used.for.moni

toring

.appl

ications

.with.no.hard.real

-time

.requ

irements,

.then.gath

ering

.seve

ral

.samp

les

in.the.node.and.tran

smitting

.them.in.one.comp

ressed

.mess

age

.can.save.much.ener

gy,

.sinc

.larg

.com-

unication

.inte

rvals

.can.be.sele

cted

.and.the.mess

age

.over

head

.is.prop

ortionally

.smal

ler

.[MDG

06].

.the.netw

ork

.has.a.tree.topol

ogy

.and.only.aggr

egated

.data.(e.g

.mean.temp

erature)

.is.need

ed,

.then.

the.info

rmation

.of.indi

vidual

.sens

ors

.(tem

perature)

.can.be.combi

ned

.and.only.the.aggr

egated

.data.can.

be.for

warded

.to.the.end.use

.[KE

W02].

.the.netw

ork

.is.used.for.moni

toring

.or.sensi

.purp

oses

.with.hard.real

-time

.requ

irements,

.then.the.

node.need

.to.wake

-up

.regu

larly

.to.sens

.chan

ges

.in.the.envi

ronment.

. e.comm

.choi

.woul

.be.

peri

odic sampling

.as.it.is.well.cove

red

.by.theo

ry.

. is.mean

.the.node.wake

.up.in.cons

tant

.inte

rvals

.for.

samp

ling

.and.tran

smitting

.each.samp

.(Fig

ure

.10.6

a).

.Howe

ver,

.real

-world

.signa

.usua

lly

.chan

.thei

dyna

mics

.over.time

. e.illu

mination

.is.one.extr

eme

.exam

ple

.as.it.can.chan

.duri

.the.day.with.mov-

.clou

.and.objec

.inst

antly,

.whil

.noth

ing

.chan

ges

.at.nigh

.when.it.is.dark

. e.freq

uent

.chan

ges

duri

.the.day.requ

ire

.a.smal

.samp

ling

.peri

od,

.whic

.is.ine

cient

.in.the.nigh

.with.only.smal

.chan

ges

and.sam

pled

.val

ues

.tha

.are.ver

.simi

lar.

Ultralow-Power Wireless Communication 10-7

Adaptive sampling.approaches.adjust.the.sampling.intervals.based.on.dierent.criteria,.e.g.,.network.

load.[GDD04],.round.trip.time,.signal.dynamics,.sampling.error,.or.operation.mode.. ey.can.be.gener-

ally

.distinguished.in.adaptive.periodic.sampling.and.event-based.sampling.approaches.

Adaptive periodic sampling

.adjusts.the.sampling.interval.of.periodic.sampling,.for.example,.in.Figure.

10.6b,.where.the.sampling.period.is.increased.aer.the.step.response.passed.its.peak.aer.6.min.. e.ben-

et

.is.that.approaches.for.periodic.systems.can.still.be.used,.e.g.,.for.signal.reconstruction.or.fault-tolerance.

mechanisms.. e.downside.is.that.reaction.times.are.limited.by.the.constant.sampling.period.

Event-based sampling

. transmits. only. signicant. changes. in. sampling. values.. Dierent. decision.

criteria.may.be.used.like.the.dierence.between.the.actual.and.last.transmitted.value.(send-on.delta.

sampling.[NK04]).or.the.integral.of.this.dierence.[VK07b]..Send-on-delta,.also.referred.as.absolute.

deadband.sampling.[OMT02],.is.the.most.common.approach..It.samples.with.a.constant.interval.T

,.but.

sends.a.sample.y(t).only.if.it.diers.more.than.a.threshold.δ.from.the.last.transmitted.sample.y(t

),.hence.

|y(t).−.y(t

)|.>.δ.. e.benet.of.these.approaches.is.that.the.number.of.transmitted.messages.is.reduced.

based.on.the.signal.dynamic..As.sampling.and.processing.usually.cost.less.energy.than.transmission,.

event-based.sampling.is.generally.more.energy.ecient..Recent.event-based.sampling.approaches.adjust.

also.the.period.of.wake-ups.to.the.signal.dynamics.[PVK09].to.remove.the.overhead.of.periodic.sam-

pling

.and.improve.energy.eciency.

Model-based reconstruction

.may.increase.the.energy.eciency.of.adaptive.sampling.approaches.even.

further.by.permitting.larger.sampling.intervals,.because.they.use.a.process.model.in.the.sender.and.

receiver.to.reconstruct.the.signal.transition.between.samples.. ey.were.applied.on.adaptive.periodic.

sampling.[MA02].and.event-based.sampling.[LYT06]..However,.they.need.precise.process.models—

otherwise

.the.sampling.quality.decreases.dramatically.

Also,

. closed-loop. controls. can. be. run. in. an. ultralow-power. network. using. adaptive. sampling.

approaches..However,.the.usage.of.standard,.nonadapted.control.algorithms.like.proportional–integral–

derivative

.(PID).lead.to.a.signicant.degradation.of.control.loop.performance.[VK07b].. e.reason.is.

that.these.control.algorithms.are.based.on.periodic.sampling.and.are.not.adapted.to.the.rare,.non-

periodic

.update.events..A.uniform.theory.for.such.event-based controls.does.not.exist.and.their.properties.

were.analyzed.only.for.some.scenarios.[A07].

.degradation.of.control.loop.performance.is.also.visible.in.Figure.10.6.. e.maximum.of.the.step.

response.for.the.event-based.sampling.in.Figure.10.6c.is.slightly.higher.with.22.7°C.than.in.Figure.

10.6a.and.b.with.22.55°C..Note.that.the.sampling.period.of.the.event-based.sampling.is.identical.

3 6 9 12

20.5

21.5

Room temperature in °C

22.5

Time in min

3 6 9 12

20.5

21.5

22.5

Time in min

(a) (b) (c)

3 6 9 12

20.5

21.5

22.5

Time in min

Constant sampling period

All samples are transmitted

Adjusts interval of periodic sampling

All samples are transmitted

Bases on a periodic sampling

May adjust sampling interval

Sends only signicant samples

Original value

Transmitted samples

Reconstructed signal

Original value

Transmitted samples

Reconstructed signal

Original value

Transmitted samples

Reconstructed signal

Energy-eciency

FIGURE 10.6 Step.responses.for.dierent.sampling.approaches.in.a.temperature.control.loop:.(a).periodic.

sampling.every.30.s,.(b).adaptive.periodic.sampling.every.30.s.for.the.rst.3.min.and.90.s.aerward,.and.(c).event-

based

.(send-on-delta).sampling.using.a.sampling.interval.of.30.s.and.a.delta.of.0.25°C.

10-8 Industrial Communication Systems

to.the.periodic.sampling.in.Figure.10.6a,.only.less.messages.are.transmitted.to.the.PID.controller..

Similar.simulations.can.be.used.during.design.time.to.analyze.specic.scenarios.[LYT06],.but.they.

are.not.always.feasible.in.practice.due.to.time.constraints..Model-based.control.approaches.are.more.

robust.as.they.estimate.the.signal.transition.between.samples.[LYT06],.[A07],.but.they.require.system.

models.that.are.oen.unknown..Heuristic.algorithms.do.not.require.such.explicit.models.and.still.

allow.an.acceptable.quality-of-control.level.[VK07a].

Figure

.10.7.compares.periodic.sampling.with.event-based.control.using.a.model-based.and.heuristic.

algorithm.for.the.step.response.shown.in.Figure.10.6.. e.control.loop.performance.(quality.of.control).

is.measured.by.the.integral.absolute.error.of.control.IAE

,.which.equals.the.integral.of.the.dierence.

between.the.set-point.and.actual.value.

.comparison.in.Figure.10.7.shows.that.with.increasing.threshold.δ,.the.error.IAE

.raises.and.

the.event.rates.N

.decrease..For.larger.threshold.values.(δ.>.0.1),.the.event.rates.do.not.decrease.sig-

nicantly,

.but.the.error.IAE

.continues.to.grow.. e.model-based.algorithm.is.less.sensitive.to.the.

threshold.value.due.to.a.more.precise.estimation.of.the.system.behavior..However,.it.requires.a.detailed.

system.model,.contrary.to.the.heuristic.approach..Both.algorithms.perform.eectively,.and.reduce.the.

number.of.events.to.up.to.one.third.of.that.of.the.equivalent.periodic.loop,.while.retaining.a.similar.

quality.of.control.

10.5 Conclusion and Open topics

Ultralow-power.wireless.communication.devices.require.a.holistic.design.from.hardware.to.appli-

cation.

. e.dierent.aspects.that.need.to.be.considered.during.design.and.common.solutions.were.

introduced.in.individual.sections.of.this.chapter..A.general.best.practice.solution.for.ultralow-power.

communication.devices.cannot.be.given,.as.minimizing.the.energy.consumption.requires.manual.

tuning.of.the.device,.protocols,.topology,.and.application.to.the.requirements.of.each.scenario..Design.

tools.that.support.engineers.in.this.task.do.not.exist.yet..Model-driven.design.approaches.for.device.

applications.can.support.the.creation.of.optimized,.energy-ecient.applications.and.communication.

protocols..Other.open.research.questions.cover.the.synchronization.of.application.and.communica-

tion

.protocols,.for.example,.the.synchronization.of.sampling.intervals.of.device.applications.with.

rendezvous-based.MAC..Especially,.cross-layer.protocols.that.consider.the.requirements.of.event-

based,

.real-time.actuation.and.combine.a.fast.delivery.of.messages.to.devices.with.a.low.duty-cycle.

are.rarely.researched..Current.research.also.covers.the.extended.usage.and.miniaturization.of.energy-

harvesting

.for.ultralow-power.devices.

0.05 0.1 0.15 0.2 0.25 0.3

100

120

140

160

180

200

Event rates N

Event-based with model-based controller

Event-based with heuristic controller

Periodic control

δ in °C(b)

0.05 0.1 0.15 0.2 0.25 0.3

δ in °C

Integral absolute error of control

IAE

in °C s

Event-based with model-based controller

Event-based with heuristic controller

Periodic control

(a)

FIGURE 10.7 Dependence.of.performance.characteristics.on.the.send-on-delta.threshold.δ..(a).Control.error.

IAE

.and.(b).event.rates.N

Ultralow-Power Wireless Communication 10-9

references

[A07].K..J..Aström,.Event.based.control,.in.Analysis and Design of Nonlinear Control Systems: In Honor of

Alberto Isidori,

.Sp

ringer

.Ver

lag,

.pp

.127–147,.Ne

.Yo

rk,

.2007.

YAH03]

.M..Bue

ttner,

.G..V..Yee,.E..And

erson,

.and.R..Han

.X-M

AC:

.A.sho

.pre

amble

.MAC.pro

tocol

for. dut

y-cycled

. wire

less

. sens

. netw

orks,

. in. Pro

ceedings of Fourth International Conference on

Embedded Networked Sensor Systems,

.pp

.307–320,.Bo

ulder,

.CO

.Oc

tober–November

.2006.

L03]

.T..V..Dam.and.K..Lan

gendoen,

.An.ada

ptive

.ene

rgy-ecient

.MAC.pro

tocol

.for.wir

eless

.sen

sor

net

works,

.in. Pro

ceedings of First ACM Conference on Embedded Networked Sensor Systems,

.Los.

Ang

eles,

.CA,.Nov

ember

.200

[EO08]

. EnOcea

. GmbH,. Energ

. for. fre

e—wireless

. tec

hnology

. wit

hout

. bat

teries,

. EnOcea

. Alli

ance,

Whit

.Pa

per,

.2008.

[FFH05]

.B..Fran

kenstein,

.K..J..Frö

hlich,

.D..Hents

chel,

.and.G..Repp

.Microsys

tem

.for.signa

.pro

cessing

app

lications,

.in.Pro

ceedings of 10th SPIE International Symposium on Nondestructive Evaluation for

Health Monitoring and Diagnostics,

.Vo

.5770,.pp

.152–159,.San.Diego

.CA,.2005.

[GDD04]

.I..A..Grava

gne,

.J..M..Dav

is,

.J..J..DaCun

ha,

.and.R..J..Mar

.II,.Ban

dwidth

.red

uction

.for.cont

roller

are

.netw

orks

.using.adap

tive

.sam

pling,

.in.Pro

ceedings of International Conference on Robotics and

Automation,

.pp

.5250–5255,.Ne

.Orle

ans,

.LA,.Ap

ril

.2004.

EW02]

.L..Kri

shnamachari,

.D..Est

rin,

.and.S..Wic

ker,

.The.imp

act

.of.dat

.agg

regation

.in.wir

eless

sen

sor

.netw

orks,

.in.Pro

ceedings of 22nd International Conference on Distributed Computing Systems

Workshops,

.pp

.575–578,.Vienn

.Au

stria,

.Ju

.2–5,.2002.

[LY

T06].F..L..Lian,.J..K..Yook,.D..M..Tilbury,.and.J..Moyne,.Network.architecture.and.communication.

mod

ules

. for. guara

nteeing

. accepta

ble

. cont

rol

. and. communic

ation

. perf

ormance

. for. netw

orked

mul

ti-agent

.syst

ems,

.IEEE Tr

ansactions on Industrial Informatics,

.2(1),.12–24,.2006.

[MA02]

. L.. A.. Mon

testruque

. and. P..J.. Ants

aklis,

. Mode

l-based

. netw

orked

. cont

rol

. system

. Neces

sary

and.sucient.condi

tions

.for.sta

bility,

.in.Pro

ceedings of 10th Mediterranean Conf. on Control and

Automation,

.Li

sbon,

.Po

rtugal,

.Ju

.2002.

[MB04]

.S..Mah

lknecht

.and.M..Boc

.CSMA-MPS:.A.minimum.pre

amble

.sam

pling

.MAC.pro

tocol

.for.

low.pow

.wire

less

.sens

.netw

orks,

.in.Pro

ceedings of the Fih International Workshop on Factory

Communication Systems,

.pp

.73–80,.Vienn

.Au

stria,

.Sep

tember

.22–24,.2004.

DG06]

.G..Mat

hur,

.P..Des

noyers,

.D..Gan

esan,

.and.P..She

noy,

.Ult

ra-low

.pow

.dat

.sto

rage

.for.sen

sor

.net-

rks,

.in.Pro

ceedings of Fih International Conference on Information Processing in Sensor Networks,

pp..374–381,.ACM,.Ne

.Yo

rk,

.2006.

[MGY04]

.P..D..Mit

cheson,

.T..C..Gre

en,

.E..M..Yea

tman,

.and.A..S..Holm

es,

.Arc

hitectures

.for.vib

ration-

driven

.micro

power

.gen

erators,

.Jo

urnal of Microelectromechanical Systems,

.13(3),.429–440,.2004.

[NK04]

.M..Neuge

bauer

.and.K..Kab

itzsch,

.A.new.pro

tocol

.for.a.low.pow

.sens

.netw

ork,

.in.Pro

ceedings

of IEEE International Conference on Performance, Computing, and Communications,

.pp..393–399,.

Phoenix,.AZ,.April.2004.

[NPK05]

.M..Neuge

bauer,

.J..Plönnigs,.and.K..Kab

itzsch,

.A.new.beaco

.order.adap

tation

.algo

rithm

.for.

IEEE.802.15.4.netw

orks,

.in.Pro

ceedings of Second European Workshop on Wireless Sensor Networks,

pp..302–311,.Is

tanbul,

.Tur

key,

.Ja

nuary–February

.2005.

PK06]

.M..Neu

gebauer,

.J..Plo

ennigs,

.and.K..Kab

itzsch,

.Eva

luation

.of.ene

rgy

.cos

.for.sin

gle

.hop.vs..mul

hop.wit

.res

pect

.to.top

ology

.par

ameters,

.in.Pro

ceedings of IEEE International Workshop on Factory

Communication Systems,

.pp..175

–182,

.Tor

ino,

.Ita

ly,

.Jun

.200

[OMT02]

.P..Ota

nez,

.J..Moy

ne,

.and.D..Til

bury,

.Using.deadb

ands

.to.red

uce

.communic

ation

.in.netw

orked

cont

rol

.system

.in.Pro

ceedings of American Control Conference,

.Vol

.4,.pp..3015–3020,.Anc

horage,

AK,.Ma

.2002.

[PH

C04]

.J..Pol

astre,

.J..Hil

.and.D..Cul

ler,

.Vers

atile

.low.pow

.medi

.access.for.wire

less

.sens

.netw

orks,

in. Pro

ceedings of Second International Conference on Embedded Networked Sensor Systems,

pp.95–107,.ACM.Press,.New.York,.2004.

10-10 Industrial Communication Systems

[PS05].J..A..Paradiso.and.T..Starner,.Energy.scavenging.for.mobile.and.wireless.electronics,.IEEE Pervasive

Computing,

.4(1),.18–27,.2005.

[PVK09]

. J.. Ploennigs,. V.. Vasyu

tynskyy,

. and. K.. Kab

itzsch,

. Com

parison

. of. energ

y-ecient

. sam

pling

meth

ods

.for.WSN

.in.buildin

.aut

omation

.scena

rios,

.in.Pro

ceedings of 14th IEEE International

Conference on Emerging Technologies and Factory Automation,

.pp..1053–1060,.Palm

.de.Mal

lorca,

Spa

in,

.Sep

tember

.22–25,.2009.

[R03]

.S..J..Roun

dy,

.Energ

.sca

venging

.for.wire

less

.sens

.nodes.wit

.a.foc

.on.vib

ration

.to.elec

tricity

conv

ersion,

.PhD.th

esis,

.Uni

versity

.of.Ca

lifornia,

.Ber

keley,

.CA,.2003.

[RSS02]

.V..Rag

hunathan,

.C..Sch

urgers,

.S..Par

.and.M..B..Sri

vastava,

.Energ

y-aware

.wire

less

.microsen

sor

netw

orks,

.IEEE Si

gnal Processing Magazine,

.19(2),.40–50,.2002.

[SD

E07]

. G.. Sch

eible,

. D..Dzung

. J..Endres

en,

. and. J.-E.. Fre

. Unp

lugged

. but.conne

cted—Design

. and.

implem

entation

.of.a.tru

.wire

less

.rea

l-time

.sens

or/actuator

.inter

face,

.IEEE In

dustrial Electronics

Magazine,

.1(2),.25–34,.2007.

[SS97]

.M..Sto

rdeur

.and.I..Sta

rk,

.Low.pow

.ther

moelectric

.genera

tor-self-sucient

.energ

.sup

ply

.for.

micro.system

.in. Pro

ceedings of 16th International Conference on ermoelectrics,

.pp..575–577,.

Dresden,.Ger

many,

.Augu

.26–29,.1997.

[ST

S02]

.C..Sch

urgers,

.V..Tsia

tsis,

.and.M..B..Sri

vastava,

.STEM:.Top

ology

.man

agement

.for.energ

.ecient.

sens

.netw

orks,

.in.Pro

ceedings of IEEE Aerospace Conference,

.Vol

.3,.pp..1099–1108,.Big.Sky

.MT,.

Mar

.2002.

[TG02]

.D..Tian.and.N..D..Georganas,.A.coverage-preserving.node.scheduling.scheme.for.large.wireless.

sens

.netw

orks,

.in.Pro

ceedings of First ACM International Workshop on Wireless Sensor Networks

and Applications,

.pp

.32–41,.Ne

.Yo

rk,

.Sep

tember

.2002.

[VK07a]

.V..Vasyu

tynskyy

.and.K..Kab

itzsch,

.Simp

.PID.cont

rol

.algo

rithm

.adap

ted

.to.deadb

and

.sam-

ling,

.in.Pro

ceedings of 12th IEEE International Conference on Emerging Technologies and Factory

Automation,

.pp

.932–940,.Pa

tras,

.Gr

eece,

.Sep

tember

.2007.

[VK07b]

. V.. Vasyu

tynskyy

. and. K.. Kab

itzsch,

. Towa

rds

. comp

arison

. of. deadb

and

. sam

pling

. types,. in.

Pro

ceedings of IEEE International Symposium on Industrial Electronics,

.pp..2899–2904,.Vigo

.Spa

in,

June.4–7,.2007.

[YHW04]

.W..Ye,.J..Heidema

nn,

.and.D..Est

rin,

.Medi

.access.cont

rol

.wit

.coor

dinated

.adap

tive

.sleep

ing

for.wir

eless

.sen

sor

.net

works,

.IEEE/ACM T

ransactions on Networking,

.12(3),.493–506,.2004.

11-1

11.1 Introduction

 is.chapter.provides.an.introduction.to.wireless.multimedia.sensor.network.(WMSN)..We.introduce.the.

WMSN.tech

nology

.by.disc

ussing

.the.WMSN.arch

itecture

.and.WMSN.hard

ware.

.WMSN.arch

itecture

outl

ines

.the.die

rent

.netw

ork

.con

gurations

.that.can.be.depl

oyed

.when.sett

ing

.up.a.WMSN.like.sing

le-tier

at,.sing

le-tier

.clus

tered,

.and.mult

itiered

.arch

itecture.

. e.next.sect

ion

.on.WMSN.hard

ware

.(Sec

tion

.11.4

)

cove

.the.wire

less

.motes.that.are.used.in.depl

oying

.the.WMSN.itse

lf,

.whic

.incl

ude

.low-

resolution

.motes

medi

um-resolution

.motes

.and.high

-resolution

.motes..Each.mote.has.addi

tional

.proc

essing

.and.sens

ing

capa

bility,

.whic

.makes.it.suit

able

.for.die

rent

.appl

ications.

. e.foll

owing

.sect

ion

.(Sec

tion

.11.5

)

.then.

descr

ibes

.pote

ntial

.WMSN.appl

ications.

. ese.appl

ications

.incl

ude

.surv

eillance,

.trac.moni

toring,

.per-

onal

.and.heal

.care

.habi

tat

.moni

toring,

.and.targ

.track

ing.

.Howe

ver,

.these.appl

ications

.brin

.with.them.

a.new.set.of.tech

nical

.chal

lenges

.like.QoS,.high.band

width,

.loca

lized

.proc

essing

.and.data.fusi

on,

.ener

Industrial Strength

Wireless Multimedia

Sensor Network

Technology

11.1. Introduction..................................................................................... 11-1

11.2

. Wirel

ess

.Sen

sor

.Net

work................................................................ 11-2

WSN.System.Requirements. •. Wireless.Multimedia.Sensor.Networks

11.3. WMSN.Architecture.......................................................................11-4

Single-Tier.Flat.Architecture. •. Single-Tier.Clustered.

Architecture. •. Multitier.Architecture

11.4. WMSN.Hardware............................................................................11-6

Low-Resolution.WMSN.Motes. •. Medium-Resolution.WMSN.

Motes. •. High-Resolution.WMSN.Motes

11.5. Applications.of.WMSNs.................................................................11-8

Surveillance. •. Trac.Monitoring. •. Personal.and.HealthCare. •. .

Habitat.Monitoring. •. Target.Tracking

11.6. WMSNs’.Technical.Challenges....................................................11-10

WMSN.Application-Specic.QoS.Requirement. •. Scalable.and.

Flexible.Architectures.and.Protocols.to.Support.Heterogeneous.

Applications. •. High.Bandwidth. •. Localized.Processing.and.

Data.Fusion. •. Energy-Ecient.Design. •. Reliability.and.Fault.

Tolerance. •. Multimedia.Coverage. •. Integration.with.IP.and.

Various.Other.Wireless.Technologies

11.7. Conclusion...................................................................................... 11-12

References................................................................................................... 11-12

Vidyasagar Potdar

Curtin University

of Tec

hnology

Atif Sharif

Curtin University

of Tec

hnology

Elizabeth Chang

Curtin University

of Tec

hnology

11-2 Industrial Communication Systems

ecient.design,.multimedia.coverage,.etc.,.which.are.then.discussed.in.depth.to.give.the.reader.complete.

understanding.of.these.needs.since.any.WMSN.application.would.need.to.keep.these.challenges.in.mind.

when.developing.new.motes.or.deploying.WMSNs.. is.concludes.the.chapter..Chapter.12.will.focus.on.

WMSN.communication.protocol.stack,.which.forms.the.key.basis.of.the.wireless.network.communication.

11.2 Wireless Sensor Network

Wireless.sensor.networks.(WSN).as.shown.in.Figure.11.1.have.attracted.wide.attention.of.research.com-

munity

.during.the.past.few.years.in.applications.ranging.from.environmental.monitoring,.heath.care,.

to.mission-critical.industrial.and.military.applications.. is.growing.interest.can.be.largely.attributed.

to.new.applications.enabled.by.large-scale.networks.of.small.devices.capable.of.harvesting.informa-

tion

.from.the.physical.environment.[AMC07],.performing.simple.processing.on.the.extracted.data.and.

transmitting.it.to.remote.locations.

WSN

.is.composed.of.intelligent.embedded.nodes.capable.of.communicating.over.wireless.link.with.

base.station.and.other.nodes..WSN.node.as.shown.in.Figure.11.2.comprises.low-power.sensing.devices,.

Server station

Node

Wireless link

Sink node

Sensor

Sensor node

FIGURE 11.1 WSN.architecture.

Sensor unit

(CO

, light,

temperature, etc.)

Central processing unit

(FPGA, microcontroller,

microprocessor, etc.)

WSN node

Power unit

(dc–dc module, battery,

filtering, etc.)

Communication unit

(RF transceiver,

bluetooth/Zigbee,

RS232 standard, etc.)

FIGURE 11.2 WSN.node.

Industrial Strength Wireless Multimedia Sensor Network Technology 11-3

embedded.processor,.communication.channel,.and.power.module.. e.embedded.processor.is.generally.

used.for.collecting.and.processing.the.signal.data.taken.from.the.sensors..Sensor.element.produces.a.

measurable.response.to.a.change.in.the.physical.condition.like.temperature,.humidity,.measuring.car-

bon

.dioxide,.etc.. e.wireless.communication.channel.provides.a.medium.to.transfer.the.information.

extracted.from.the.sensor.node.to.the.exterior.world,.which.may.be.a.computer.network,.and.inter-node.

communication..Finally,.power.unit.mostly.comprises.AA-size.batteries.and.dc–dc.module.for.feeding.

the.entire.WSN.node.

11.2.1 WSN System requirements

Here,.we.discuss.some.of.the.characteristic.requirements.of.a.system.comprising.wireless.sensor.nodes..

 e.system.should.be

. 1..Fault tolerant:. e.system.should.be.robust.against.node.failure.(running.out.of.energy,.physical.

destruction,.H/W,.S/W.issues,.etc.).

. 2.. Scalable:. e. system. should. support. a. large. number. of. sensor. nodes. to. cater. for. dierent.

applications.

. 3..Long life:. e.node’s.lifetime.entirely.denes.the.network’s.lifetime.and.it.should.be.high.enough..

 e.sensor.node.should.be.power.ecient.against.the.limited.power.resource.that.it.has.since.it.is.

dicult.to.replace.or.recharge.thousands.of.nodes.. e.node’s.communication,.computing,.sens-

ing,

.and.actuating.operations.should.be.energy.ecient.too.

. 4.. Programmable:. e.reprogramming.of.sensor.nodes.in.the.eld.might.be.necessary.to.improve.

exibility.

. 5..Secure:. e.system.should.have.the.following:

. a.. Access control:.To.prevent.unauthorized.attempts.

. b.. Message integrity:.To.detect.and.prevent.unauthorized.changes.in.the.message.

. c.. Condentiality:.To.assure.that.sensor.node.should.encrypt.messages.so.that.only.those.nodes.

that.have.the.secret.key.would.listen.

. d.. Replay protection:.To.assure.that.sensor.node.should.provide.protection.against.an.adver-

sary

.reusing.an.authentic.packet.for.gaining.condence/network.access;.man-in-the-middle.

attack.can.be.prevented.by.time-stamping.the.data.packets.

. 6.. Aordable:. e.system.should.use.low-cost.devices.since.the.network.comprises.thousands.of.

sensor.nodes.. e.installation.and.maintenance.of.system.elements.should.also.be.signicantly.

low.to.make.its.deployment.realistic.

11.2.2 Wireless Multimedia Sensor Networks

Wireless.sensor.network.node,.if.equipped.with.audio-.or.video-sensing.device,.forms.a.WMSN.node.as.

shown.in.Figure.11.3..And.the.network.formed.by.such.nodes.(WMSN).is.dened.as.a.network.of.wire-

lessly

.interconnected.devices.that.are.able.to.ubiquitously.retrieve.multimedia.content.such.as.video.and.

audio.streams,.still.images,.and.scalar.sensor.data.from.the.environment.. e.key.application.in.which.

WMSN.is.widely.used.includes.surveillance.systems,.intelligent.trac.management.systems,.habitat.

monitoring,.and.military.applications..All.these.applications.are.heterogeneous.by.nature.as.they.not.

only.require.scalar.information.but.also.multimedia.information.as.well..Multitier.WMSN.architec-

tures

.are.generally.used.for.these.applications.to.have.a.complete.understanding.of.the.environment.

WSN

Multimedia

(video + audio)

WMSN

FIGURE 11.3 Wireless.multimedia.sensor.network.evolution.from.WSN.

11-4 Industrial Communication Systems

behavior..For.such.heterogeneous.applications,.WSN.node.does.not.fulll.the.application.requirements.

beca

use

.of.the.fol

lowing

.key.asp

ects

.tha

.WMS

.nod

.hav

•

. Proc

essing power:

. Sinc

. the. mult

imedia

. info

rmation

. proc

essing

. is. comp

utationally

. inte

nsive,

high

-end

.proc

essor

.or.eld

-programmable

.gate.arra

.(FPG

A)/ASIC

.is.gene

rally

.used.for.WMSN.

node.as.opp

osed

.to.simp

.mic

rocontrollers

.as.in.the.cas

.of.WSN

•

. e.amo

unt of data

.in.WMS

.is.mor

.in.mag

nitude

.tha

.in.WSN

•

. Stor

age memory:

.Mult

imedia

.info

rmation

.loca

.proc

essing

.does.requ

ire

.temp

orary

.stor

age/buer-

ing

.of.the.data.duri

.mani

pulation

.and.duri

.sensi

ng.

.Howe

ver,

.ther

.is.no.need.of.inpu

.bue

and.high.memo

.requ

irement

.in.case.of.WSN.that.deal

.with.only.scala

.info

rmation.

•

. Communication standard which it uses for sending and receiving the multimedia and scalar information:.

UWB.is.gene

rally

.used.as.wire

less

.comm

unication

.stan

dard

.for.mult

imedia

.infor

mation

.comm

uni-

cation;

.wher

eas

.in.the.case.of.WSN,.Zigb

.is.gene

rally

.used.as.a.wire

less

.comm

unication

.stan

dard

havi

.a.data.rate.of.250.kbps

•

. Netw

ork conguration:

.As.oppo

sed

.to.star

.clus

ter,

.or.Mesh.topol

ogies

.in.WSN,.in.WMSN

.the.

basic.netw

ork

.topol

ogy,

.alth

ough

.havi

.the.same.con

guration

.stan

dard,

.is.broa

dly

.clas

sied

.as.

singl

e-tier

.or.mult

itier,

.base

.on.the.natu

.of.appl

ication

.(det

ails

.can.be.foun

.in.the.foll

owing

Sect

ion

.11.

3).

•

. Homo

geneous vs. heterogeneous:

.WSN.is.capa

ble

.of.sensi

.only.scal

.info

rmation

.(at.homo

ge-

neous

.arch

itecture;

.each.node.comp

rising

.same.sens

.and.havi

.same.comp

utational

.and.com-

unication

.powe

r);

.howe

ver,

.WMSN

.havi

.mult

itier

.con

guration,

.has.hete

rogeneous

.sensi

elem

ents

.(sc

alar

.sen

sors,

.aud

.sen

sors,

.low

/medium/high

.res

olution

.vid

.sen

sors,

.etc

.).

•

. Becau

.of.the.die

rence

.betw

een

.the.appl

ication

.envi

ronments

.and.data.feat

ures

.of.stre

.data.in.

WMSN.and.scala

.data.in.WSN,.trad

itional

.secure.Rout

ing

.for.scala

.data.is.not.t.for.stre

.data

•

. e.trad

itional

.WSN.tran

sport

.proto

col

.is.not.used.in.WMSN.beca

use

.it.is.unab

.to.meet.the.

mult

imedia

.appl

ication-specic

.QoS.para

meter.

.In.cont

rast

.to.trad

itional

.WSN.tran

sport

.proto

cols

.are.the.WMSN.proto

cols

.like.cong

estion

.cont

rol,

.end-

to-end

.reli

ability

.of.comm

unication,

and.pac

ket

.reo

rdering

.due.to.mul

tipath.

•

. e.exis

ting

.WSN.mote

.like.Telo

s/TelosB,

.Mica

.mote.fami

ly,

.did.not.supp

ort

.the.in-n

etwork

processing feature,

.which.is.generally.desirable.in.multimedia.applications.due.to.the.large.size.of.

the.mult

imedia

.info

rmation

.(aud

io/video

.stre

.or.snap

shot).

.If.such.mote

.are.used.in.mult

ime-

dia

.appl

ications,

.they.will.rapi

dly

.depl

ete

.thei

.ener

.beca

use

.of.enor

mous

.powe

.cons

umption

due.to.the.mul

timedia

.pro

cessing

.and.rad

.com

munication

.(ra

.mul

timedia

.dat

a).

11.3 WMSN architecture

Multimedia.sensor.networks,.in.particular,.exploit.multilevel.tier.architectures.for.communication.and.

netw

ork

.mana

gement.

. is.is.beca

use,

.in.addi

tion

.to.sens

.comp

ression,

.data.aggr

egation

.amon

.node

in.a.comm

.neig

hborhood

.is.nece

ssary

.for.scal

able

.deci

sion-making

.and.netw

ork

.oper

ation.

.Figu

11.3.[AMC

07,BFNPV07]

.prov

ides

.an.exam

ple

.in.whic

.hier

archical

.thre

e-tier

.arch

itecture

.[BP0

.is.

used.for.a.vide

o-based

.sens

.netw

ork.

. e.arch

itecture

.in.Figu

.11.4.depi

cts

.the.thre

.die

rent

.type

of.sen

sor

.clo

uds

.bas

.on.the.nat

ure

.of.app

lication.

11.3.1 Single-tier Flat architecture

 e.sensor.cloud.on.the.le.is.the.single-tier.at.architecture.comprising.homogeneous.sensors.(in.

fact

.same.sens

ors,

.i.e.,.havi

.the.same.sensi

ng,

.proc

essing,

.and.comm

unication

.capa

bility).

.As.we.

have.seen

.ther

.are.few.mult

imedia

.hubs.and.vide

.sens

ors

.in.it.. e.mult

imedia

.proc

essing

.hubs.have.

more.comp

uting

.powe

.as.comp

ared

.to.the.vide

.sens

.node

. e.mult

imedia

.data

.in.this.netw

ork,

are.tran

sferred

.from.the.pare

.node.to.sink/

storage

.devi

.via.the.gate

way

.foll

owing

.the.hop-b

y-hop