EAST:Energy-efficient Adaptive Scheme for …njavaid.com/Tahir-EAST.pdfEAST:Energy-efficient Adaptive Scheme for Transmission in Wireless Sensor Networks EAST:Energy-efficient Adaptive

EAST:Energy-efficient Adaptive Scheme for Transmission in Wireless Sensor Networks

EAST:Energy-efficient Adaptive Scheme for

Transmission in Wireless Sensor Networks

Muhammad Tahir

Department of Electrical Engineering

COMSATS Institute of Information Technology

Islamabad, Pakistan

[email protected]

February 02, 2013

1 / 23


Outline

Outline

Introduction to the Problem

Related Work and Motivation

Proposed Energy Efficient Adaptive Transmission Scheme

Block Diagram of EAST

EAST Flow Chart

Parameters

Simulation Results

Conclusion

2 / 23




Challenge in design of Wireless Sensor Networks (WSNs) is toreduce energy consumption of sensor nodes to prolong lifetimeof Network

Limited battery requires low power sensing, processing andcommunication system

In WSNs, sensor nodes are widely deployed in differentenvironments to collect data

Low power wireless link causes link quality variation due toenvironmental dynamics like temperature, humidity etc

Therefore, while maintaining good link quality with nodes weneed to reduce energy consumption

3 / 23




Existing schemes set some minimum transmitter power levelfor maintaining reliability

To adjust transmitter power, reference node periodicallybroadcasts a beacon message

When nodes hear a beacon message from a reference node,these nodes transmit an ACK message

Through this interaction, reference node estimate connectivitywith nodes

In Local Mean Algorithm (LMA), a reference node broadcastsLifeMsg message

4 / 23




Nodes transmit LifeAckMsg after they receive LifeMsg

Reference nodes count number of LifeAckMsgs andtransmission power is controlled by maintaining appropriateconnectivity

For example if number of LifeAckMsgs is less thanNodeMinThresh transmission power is increased

In contrast, if number of LifeAckMsgs is more thanNodeMaxThresh transmission power is decreased

Local Information No Topology/Local Information Link-stateTopology (LINT/LILT) and Dynamic Transmission PowerControl (DTPC) uses transmission power loss RSSIloss toestimate transmitter power level

5 / 23




Transmission power also controlled by Packet Reception Ratio(PRR) metric

Since RSSIloss is directly proportional to temperature.Adaptive Transmission Power Control (ATPC) adjuststransmission power dynamically according to spatial andtemporal effects

In adapting link quality for environments where temperaturevariation occur, packet overhead for transmission powercontrol should be minimized. Reducing number of controlpackets while maintaining reliability is also an importanttechnical issue

6 / 23



Proposed Energy Efficient Adaptive Transmission

Scheme

Propose Energy-efficient Adaptive Scheme for Transmission(EAST) of data in WSN,s is IEEE 802.15.4 standardcompliant

In this approach, Open-loop for temperature-aware link qualityestimation and compensation and Closed-loop feedbackprocess for

a)Logical division of network into three regions

b) Minimization of control packets overhead

Threshold transmitter power loss (RSSIloss) for each regionhelps to adapt transmitter power according to

a)Link quality changes due to temperature variation

b)Current number of nodes in that region

7 / 23



Proposed Energy Efficient Adaptive Transmission

Scheme

By adopting both open-loop and closed-loop feedbackprocesses we can divide network into three regions on thebasis of threshold RSSIloss for each region

1: A for High RSSIloss

2: B for Medium RSSIloss

3: C for Low RSSIloss

EAST has two phases, i.e., initial and run-time phases

In the initial phase, reference node builds a model for nodes ofeach region

In the run-time phases, based on the previous model, EASTadapts the link quality to dynamically maintain each link overtime

8 / 23



Mathematical Formulation of Proposed Scheme

Transmission power loss due to temperature variationformulated using relationship between RSSIloss andtemperature experimented in Bannister et al:

RSSIloss [dBm] = 0.1996 ∗ (T [C o ]− 25[C o ]) (1)

To compensate RSSIloss , relationship for required power levelis given in Eq.(2) using least square approximation:

Plevel [dBm] = [(RSSIloss + 40)/12]2.91 (2)

To compensate path loss due to distance between sensornodes free space model help to estimate actual requiredtransmitter power as given in Eq.(3):

Pt [dBm] = [η∗(Eb/N0)∗mkTB∗(4πd/λ)2+RNF ]+Plevel (3)

9 / 23


Block Diagram

Block Diagram

Power

ControllerEAST Network

Temperature

Open Loop

Closed Loop

+/-Nd(t) Nc(t)

10 / 23


EAST Flow Chart

EAST Flow Chart

START

Are Temperature Changes Detected in

Neighbour Nodes?

1:High RSSI_loss (A)

2:Medium RSSI_loss (B)

3:Low RSSI_loss (C)

1:RSSI_loss (A)

2:P_level (A)

3:Count N_A

4:N_A Desired

1:RSSI_loss (B)

2:P_level (B)

3:Count N_B

4:N_B Desired

1:RSSI_loss (C)

2:P_level (C)

3:Count N_C

4:N_C Desired

Define Threshold RSSI_loss

(A,B,C)

Broadcast

1:P_level_new(A,B,C)

2:P_save(A,B,C)

END

Yes

Keep Current

Transmitter Power

Level

No

RSSI_loss_Threshold (A,B,C)<=RSSI_loss (A,B,C)RSSI_loss_Threshold (A,B,C)>RSSI_loss (A,B,C)

N_current>=N_desired N_current<N_desired

RSSI_loss_new

(A,B,C)

=RSSI_loss_Thresh

old

RSSI_loss_new

(A,B,C)=RSSI_loss

(A,B,C)

Set the

parameters

(N,d,T)

Estimate

1:RSSI_loss

2:P_level

11 / 23


Parameters

Simulation parameters

Rounds 1200

Temperature -10-53 C 0

Distance (1-100)m

Nodes 100

Regions A,B,C

η 0.0029

SNR 0.20dB

Bandwidth 83.5MHz

Frequency 2.45GHz

RNF 5dB

Eb/N0 8.3dB

12 / 23


Parameters

Estimated parameters

Desired Nodes (A,B,C) 41,25,19

Current Nodes (A,B,C) 41,22,17

Threshold power level (A,B,C) 43.24,31.77,22.21 dBm

Nodes above threshold (A,B,C) 23,11,8

Nodes below threshold (A,B,C) 18,11,9

PRR (A,B,C) (80-98),(70-96),(63-97) %

Threshold RSSIloss ( A,B,C) 3.78,-0.61,-5.17 dBm

13 / 23


Simulation Results

Temperature for different sensor nodes

Suppose we have 100 nodes in 100*100 m2 square region andtemperature can have values in range (-10 - 53)C o for givenmeteorological condition

Each sensor node placed at random location in given area andwe clearly see variation of temperature for different nodes inWSN

10 20 30 40 50 60 70 80 90 100−10

0

10

20

30

40

50

60

Nodes (N)

Tem

pera

ture

(C

o)

Temperature (Co)

14 / 23


Simulation Results

Transmission power loss for different sensor nodes

Figure given below shows transmission power loss due totemperature variation in any environment

RSSIloss(dBm) high means that sensor node placed in regionwhere temperature is high so link not have good quality

10 20 30 40 50 60 70 80 90 100−8

−6

−4

−2

0

2

4

6

Nodes (N)

RS

SI−

loss

(dB

m)

RSSI−loss (dBm)

15 / 23


Simulation Results

Transmitter power level for different sensor nodes

After estimating RSSIloss for each node in WSN we computecorresponding transmitter power level to compensate RSSIloss

Plevel assigned to each node on basis of nodes estimatedRSSIloss

10 20 30 40 50 60 70 80 90 10015

20

25

30

35

40

45

50

Nodes (N)

Pow

er le

vel (

dBm

)

Power level (dBm)

16 / 23


Simulation Results

Transmitter power level for three region before EAST

For region A required power level high then both other regionthat shows for that region RSSIloss is large

For region B required power level is between both region Aand C and for C region required power level is less then bothother two regions

100 200 300 400 500 600 700 800 900 1000 1100 120020

25

30

35

40

45

Rounds

Pow

er le

vel(A

,B,C

) dB

m

Power level(A,B,C) dBm

17 / 23


Simulation Results

Transmitter power level for three region after EAST

After applying our propose technique we see what power levelrequired for each region

We clearly see difference between Plevel as shown in Figurethat required power level decrease for each region and forregion A it decreases maximum

100 200 300 400 500 600 700 800 900 1000 1100 120020

25

30

35

40

45

Rounds

Pow

er le

vel n

ew(A

,B,C

) dB

m

Power level new(A,B,C) dBm

18 / 23


Simulation Results

Transmitter power save for region A

100 200 300 400 500 600 700 800 900 1000 1100 12000

0.5

1

1.5

2

2.5

Rounds

Pow

er le

vel s

ave(

A)

dBm

Power level save(A) dBm

19 / 23


Simulation Results

Transmitter power save for region B

100 200 300 400 500 600 700 800 900 1000 1100 12000

0.5

1

1.5

2

2.5

Rounds

Pow

er le

vel s

ave(

B)

dBm

Power level save(B) dBm

20 / 23


Simulation Results

Transmitter power save for region C

100 200 300 400 500 600 700 800 900 1000 1100 12000

0.2

0.4

0.6

0.8

1

1.2

1.4

Rounds

Pow

er le

vel s

ave(

C)

dBm

Power level save(C) dBm

21 / 23


Conclusion

Conclusion

In this thesis , I have presented propose technique EAST tostudy temperature effect on wireless link qualityRelationship between RSSIloss and temperature has beenanalyzed for propose schemeThis scheme uses open-loop control to compensate forchanges of link quality according to temperature variationFurther extension of this scheme by dividing network intothree regions on basis of Threshold RSSIloss and assign powerlevel to each node in three regions on basis of current numberof nodes and desired number of nodes help to adapttransmitter power according to link quality variation andincrease network lifetimeCombining both open-loop temperature-aware compensationand close-loop feedback control cause significant reductionoverhead of transmission power control in a WSN

22 / 23


Questions

Questions

Thank you!

23 / 23

Documents

EAST:Energy-efficient Adaptive Scheme for …njavaid.com/Tahir-EAST.pdfEAST:Energy-efficient Adaptive Scheme for Transmission in Wireless Sensor Networks EAST:Energy-efficient Adaptive