Balanced- -energy Sleep energy Sleep Balanced Scheduling Scheme - - PowerPoint PPT Presentation

Balanced Balanced-

• energy Sleep

energy Sleep Scheduling Scheme for High Scheduling Scheme for High Density Cluster Density Cluster-

• based Sensor

based Sensor Networks Networks

• J. Deng
• J. Deng*

*, Y. Han

, Y. Han**

**, W. Heinzelman

, W. Heinzelman†

† and P.

and P. Varshney Varshney*

* * *Syracuse University

Syracuse University

** **National Taipei University, Taiwan

National Taipei University, Taiwan

† †University of Rochester

University of Rochester

Motivation Motivation

• Consider:

Consider:

• Sensor network with randomly distributed

Sensor network with randomly distributed sensors sensors

• Application: provide coverage of area for

Application: provide coverage of area for surveillance (QoS) surveillance (QoS)

• Assumption:

Assumption:

• Sensor density is higher than necessary for

Sensor density is higher than necessary for meeting QoS meeting QoS

Motivation (cont.) Motivation (cont.)

• Characteristics of sensor networks

Characteristics of sensor networks

• Low energy

Low energy

• Low bandwidth

Low bandwidth

• Networks expected to last for months unattended

Networks expected to last for months unattended

• Energy

Energy-

• efficiency is crucial

efficiency is crucial

• Exploit redundancy by powering down

Exploit redundancy by powering down unnecessary sensors unnecessary sensors

• Save energy for later when nodes are more important

Save energy for later when nodes are more important

• Sleep Scheduling Problem: Which sensors to

Sleep Scheduling Problem: Which sensors to power down? power down?

Cluster Cluster-

• based Networks

based Networks

• Base station cannot manage sensors

Base station cannot manage sensors directly directly

• Clustering provides framework for

Clustering provides framework for

• Local control

Local control

• Resource management

Resource management

• Channel access

Channel access

• Data fusion

Data fusion

• Within a cluster, how to set nodes to sleep?

Within a cluster, how to set nodes to sleep?

Assumptions Assumptions

• Dense, static, circular clusters

Dense, static, circular clusters

• Variable transmission power to reach cluster head

Variable transmission power to reach cluster head

• x

x = distance from sensor to cluster head = distance from sensor to cluster head

• Nodes distributed as 2D Poisson point process

Nodes distributed as 2D Poisson point process

• Energy savings is expected energy consumption

Energy savings is expected energy consumption were the node awake were the node awake

2 min 1

)] , [max( ) ( k x x k x Eactive + ⋅ ⋅ =

γ

λ

Initial Scheduling Techniques Initial Scheduling Techniques

• Randomized scheduling (RS)

Randomized scheduling (RS)

• Randomly put sensors to sleep

Randomly put sensors to sleep

• Each sensor sleeps with probability

Each sensor sleeps with probability

• Distance

Distance-

• based scheduling (DS)

based scheduling (DS)

• Probability

Probability p p linearly related to linearly related to x x

• Sensors further from cluster head put to sleep with

Sensors further from cluster head put to sleep with higher probability higher probability

R x R x x p

s

≤ ≤ = 2 3 ) ( β

1 < =

s

p β

Coefficient of Variation Coefficient of Variation

• Analytically determine coefficient of variation of

Analytically determine coefficient of variation of energy consumption for RS and DS energy consumption for RS and DS

2 2 γ γ 100 m 100 m R R 25, 50, 100 25, 50, 100 pkt pkt/sec /sec λ λ 10 m 10 m x xmin

min

0.1 J/sec 0.1 J/sec k k2

2

10 10-

• 6

6 J/pkt/m

J/pkt/m2

2

k k1

1

500 500 # nodes # nodes

Balanced Balanced-

• Energy Scheduling (BS)

Energy Scheduling (BS)

• p(x

p(x) ) chosen so nodes consume same amount of chosen so nodes consume same amount of energy, on average energy, on average

• Let

Let E EBS

BS(x

(x) ) be expected energy consumption of a be expected energy consumption of a node at distance node at distance x x from cluster head from cluster head

• Find

Find p(x p(x) ) such that such that E EBS

BS(x

(x) ) does not depend on does not depend on x x

• Can only energy balance certain portion

Can only energy balance certain portion β

βb

b of nodes

• f nodes
• Nodes close to cluster head not energy balanced

Nodes close to cluster head not energy balanced

R x x E x E x p x E

b b BS active BS

≤ ≤ = − =

) (

) ( )] ( 1 [ ) (

Results Results

Performance Evaluation Performance Evaluation

• Analytically

Analytically determine determine expected energy expected energy consumption consumption

• λ

λ = 100 = 100 pkts/s pkts/s

Performance Evaluation (cont.) Performance Evaluation (cont.)

• BS achieves goal of lower coefficient of variation

BS achieves goal of lower coefficient of variation

Network Lifetime Network Lifetime

• T(

T(β βd

d)

) = time when = time when β βd

d fraction of sensors run out

fraction of sensors run out

• f energy
• f energy
• Initial sensor energy =

Initial sensor energy = Ψ Ψ

• For BS,

For BS, β βb

b fraction of nodes consume same energy

fraction of nodes consume same energy

• When

When β βd

d =

= β βb

b

• In RS, nodes farther away consume more energy

In RS, nodes farther away consume more energy

• Run out of energy faster than closer nodes

Run out of energy faster than closer nodes

• In DS, network lifetime calculated numerically

In DS, network lifetime calculated numerically

) (

) (

b BS d BS

E T Ψ = β

Lifetime Results Lifetime Results

Lifetime Results (cont.) Lifetime Results (cont.)

• BS has 70% (150%) longer lifetime than

BS has 70% (150%) longer lifetime than DS (RS) for DS (RS) for β βd

d = 0.1

= 0.1

• BS has better lifetime than DS and RS for

BS has better lifetime than DS and RS for all points except all points except β βd

d = 0.5 and

= 0.5 and β βs

s < 0.27

< 0.27

• Small

Small β βs

s

fewer sensors energy balanced fewer sensors energy balanced

• 50% sensors run out of energy quickly

50% sensors run out of energy quickly

Sensing Coverage Sensing Coverage

RS DS BS

Sensing Coverage Distribution Sensing Coverage Distribution

Initial sensing coverage distribution Sensing coverage distribution after 40% nodes run out of energy

50% Sensors Remaining 50% Sensors Remaining

RS DS BS

Conclusions Conclusions

• Sleep scheduling important to extend network

Sleep scheduling important to extend network lifetime lifetime

• Balanced

Balanced-

• energy scheduling effective in

energy scheduling effective in extending lifetime while maintaining coverage extending lifetime while maintaining coverage

• Future work

Future work

• Explore different initial energies

Explore different initial energies

• Dynamically changing clusters and cluster heads to

Dynamically changing clusters and cluster heads to balance energy among all nodes balance energy among all nodes