+ All Categories
Home > Documents > Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple...

Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple...

Date post: 13-Jan-2016
Category:
Upload: dayna-arnold
View: 216 times
Download: 0 times
Share this document with a friend
Popular Tags:
11
Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities
Transcript
Page 1: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Ben GaudetteCSE 535: Mobile Computing

12/1/2010

Energy Management in WSN’s with Multiple Modalities

Page 2: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Node Characterization MSP430

Flash Read – 2.3 mA Flash Write – 13.3mA Computation – 6.33mA Frequency – 16 MHz Operation voltage – 3 V

CC2500 Receive – 13.3 mA - 2.6 ms Transmit – 0.840 ms

Sensing - PhotoElectric Base Range – 2 meters Base Energy draw – 35 mA

– 0.5 ns – 5 second refresh rate

Relationship – Quadratic E = 3.6*10^-6*R2

Page 3: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Data StructuresMatrix

BDD

S1,R1

S1,R2

… S1,RP

… SN,RP

T1 1 1 … 1 … 0

T2 0 0 … 1 … 1

… … … … … … ....

TM 0 0 … 0 … 1

BDDComputation – O(N*P)Space – O(22N*P-1)*

MatrixComputation –

O((M*N*P))Space – O((M*N*P))

Page 4: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Centralized Algorithms M3C updated

M3C (XR,Xon, S)

(1)XM3C XR;(2)S SORT_BY_WEIGHT(S);(3)if Xon = NULL(4) i 0;(5)else (6) i

MIN_WEIGHT_NODE_INDEX(Xon)

(7) XM3C ∀(XM3C, s1…si);

(8)if XM3C is ø

(9) return Xon;(10) foreach j (i+1 to length

of array S)(11) result ∀(XM3C,

sj,r, ~Si,rk)(12) if result ∉ ø(13) XM3C

result;(14) return XM3C;

Centralized Switch New

M3C_Sequential_Schedule(XR, S)

(1)T 0; Xon NULL;

(2)while XR ∉ ø

(3) Xon M3C(XR, Xon, S);

(4) Δk t;

(5) T T + Δk;

(6) UPDATE_WEIGHTS(S, Xon, Δk, T);

(7) Sd ø;(8) foreach s ∈ S(9) foreach range ∈ s(10) if ws,r(T) < t

(11) Sd Sd ⋃ sr

(12) if Sd = ø

(13) XR A(XR, Sd);

(14) return T;

Page 5: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Distributed Algorithm– Neighbor Discovery Phase• Share information with neighbors.• Predicted lifetime, each available

range and the targets the ranges can cover

– Arbitration Phase• Starting with the weakest node.• Check to see if any points

exclusive to the largest range can be covered by stronger nodes.

• If no turn on to that range, else repeat for lower ranges.

Page 6: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Results - SetupTests

Simulations – Based off of some of the power numbers

TargetsVaried between 3 and 5Grid points are targetsBase Station in the middle

Ranges: 1-3{1.5, 3, 4.5}, {1.5, 3}, {1.5}

Sensors 20 and 50.Random LocationsSame starting Energy

Performed each test 10 times.

Page 7: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Results – Number of Targets

20 Sensors

50 Sensors

Results – Number of Ranges

3 x 3 Grid

5 x 5 Grid

Page 8: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

Results – Number of Sensors

4x4 Grid of Targets, 3 Ranges

Page 9: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

ConclusionResults

Data Structure – Speed vs. MemoryAlgorithms – Multi-range is beneficial in sparse

environments.Outperforms AR-SC by an average factor of 1.8.When many sensors are close to one another (densely

placed), the algorithms show no benefit of having multiple ranges.

Future WorkDistributed Algorithms implementation and data

structure analysis?Improve the simulator

Battery modelMemory access power predictionVoltage scaling

Expand the scope of the problemEnergy Harvesting

Page 10: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

QUESTIONS?

Page 11: Ben Gaudette CSE 535: Mobile Computing 12/1/2010 Energy Management in WSN’s with Multiple Modalities.

ReferencesS. Dasika, S. Vrudhula, K. Chopra, S.

Ramasubramanian, A framework for battery-aware sensor management, in: Proc. of the Conf. on Design, Automation, and Test in Europe (DATE), 2004, pp. 1–6.

M. Cardei, J. Wu, M. Lu, and M. O. Pervaiz, Maximum networklifetime in wireless sensor networks with adjustablesensing ranges. Proc. of IEEE International Conference onWireless and Mobile Computing, Networking and Communications(WiMob), 2005.

http://focus.ti.com/docs/toolsw/folders/print/ez430-rf2500.html


Recommended