Sunteți pe pagina 1din 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

and Related Works Proposed Method Simulation Conclusion Reducing The Data Transmission in Wireless Sensor Networks

Reducing The Data Transmission in Wireless Sensor Networks Using The Principal Component Analysis

.

A. Rooshenas, H. R. Rabiee, A. Movaghar, M. Y. Naderi

The Sixth International Conference on Intelligent Sensors, Sensor Networks and Information Processing ISSNIP 2010

DML

8 Dec 2010

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 1
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
1 / 34

DML

.

.

.

.

Outline Introduction and Related Works Proposed Method Simulation Conclusion

and Related Works Proposed Method Simulation Conclusion . 1 . Introduction and Related Works Background PCAg

.1 .

Introduction and Related Works

Background PCAg DPCA

.

.2

Proposed Method

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

.

.3

Simulation

Settings

Results

.

.4

Conclusion

Future Works References Appendix

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 2
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
2 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Problem Definition

PCAg Simulation Conclusion DPCA Problem Definition Different Power Consumption in Routing Tree Leaf Nodes

Different Power Consumption in Routing Tree

Leaf Nodes

Intermediate Node

DML

.

Problems Intermediate nodes should relay other packets to the base station.

Lifetime of intermediate nodes affect the longevity of Wireless Sensor Networks

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 3
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
3 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Dimension Reduction

PCAg Simulation Conclusion DPCA Dimension Reduction DML . x 2 .v . .x 1 8 Dec

DML

.x 2

.v . .x 1 8 Dec 2010 Amirmohammad Rooshenas
.v
.
.x 1
8 Dec 2010
Amirmohammad Rooshenas
. . . . . . Reducing The Data Trans. in WSN using PCA 4
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
4 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Dimension Reduction

PCAg Simulation Conclusion DPCA Dimension Reduction DML . x 2 .v . . . x 1

DML

.x 2

.v . .
.v
.
.

.x 1

.y = v T x

Reduction DML . x 2 .v . . . x 1 . y = v T
.v . . . . . .
.v
.
.
.
.
.
.

8 Dec 2010

Amirmohammad Rooshenas

Reducing The Data Trans. in WSN using PCA

5 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Dimension Reduction

.x 2

Simulation Conclusion DPCA Dimension Reduction . x 2 .v .y = v T x .v .ˆx
.v .y = v T x .v .ˆx = vy . .x 1 . .
.v
.y = v T x
.v
.ˆx = vy
.
.x 1
.
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
6 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Related Works

PCAg Simulation Conclusion DPCA Related Works Principal Component Aggregation (PCAg). Unsupervised

Principal Component Aggregation (PCAg).

Unsupervised

Supervised

Distributed Principal Component Analysis (DPCA).

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 7
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
7 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Proposed Method PCAg Simulation Conclusion DPCA Principal Component Aggregation Base station gathers

Principal Component Aggregation

Base station gathers observations from sensor networks

.

.BS .x 1 , x 2 , x 3 , x 4 , x 5
.BS
.x 1 , x 2 , x 3 , x 4 , x 5
.x 6 , x 7 , x 8 , x 9
.n 3
.n 6
.n 9
.x 7 , x 8 , x 9
.x 1 , x 2
.x 1 , x 2 , x 3
.x 4 , x 5
.x 7 , x 8
.n 2
.n 5
.n 8
.x 1
.x 4
.x 7
.n 1
.n 4
.n 7
.
.
.
.
.
.

DML

8 Dec 2010

Amirmohammad Rooshenas

Reducing The Data Trans. in WSN using PCA

8 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Proposed Method PCAg Simulation Conclusion DPCA Principal Component Aggregation Base station calculates

Principal Component Aggregation

Base station calculates eigenvectors and sends them back to network

.

.BS .v 1 , v 2 , v 3 , v 4 .v 5 ,
.BS
.v 1 , v 2 , v 3 , v 4
.v 5 , v 6 , v 7 , v 8 , v 9
.n 3
.n 6
.n 9
.v 1 , v 2
.v 1 , v 2 , v 3
.v 4 , v 5
.v 7 , v 8 , v 9
.v 7 , v 8
.n 2
.n 5
.n 8
.v 1
.v 4
.v 7
.n 1
.n 4
.n 7
.
.
.
.
.
.

DML

8 Dec 2010

Amirmohammad Rooshenas

Reducing The Data Trans. in WSN using PCA

9 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Proposed Method PCAg Simulation Conclusion DPCA Principal Component Aggregation Sensor nodes send partial

Principal Component Aggregation

Sensor nodes send partial PCs in each epoch

The base station reconstructs observations

Reconstruction error is not available at the base station

The base station calculates eigenvectors periodically

.

.

.BS .y = ∑ 9 i=1 v i ∗ x i ∑ 3 . i=1
.BS
.y = ∑ 9
i=1 v i ∗ x i
3
.
i=1 v i ∗ x i
.n 3
.n 6
.n 9
9
.
i=7 v i ∗ x i
2
5
8
.
.
i=1 v i ∗ x i
i=4 v i ∗ x i
i=7 v i ∗ x i
.n 2
.n 5
.n 8
.v 1 ∗ x 1
.v 4 ∗ x 4
.v 7 ∗ x 7
.n 1
.n 4
.n 7
.
.
.
.
.
.

DML

8 Dec 2010

Amirmohammad Rooshenas

Reducing The Data Trans. in WSN using PCA

10 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Proposed Method PCAg Simulation Conclusion DPCA Principal Component Aggregation The base station sends PCs

Principal Component Aggregation

The base station sends PCs to the network

Each sensor can calculate its reconstruction error

e i = x i y v i

Updating one observation may change all other elements of an eigenvector

Update decision is ambiguous

DML

8 Dec 2010

.

Amirmohammad Rooshenas

.BS .y .n 3 .n 6 .n 9 .y .y .y .y .y .n 2
.BS
.y
.n 3
.n 6
.n 9
.y
.y
.y
.y
.y
.n 2
.n 5
.n 8
.y
.y
.y
.n 1
.n 4
.n 7
.
.
.
.
.
.

Reducing The Data Trans. in WSN using PCA

11 / 34

Outline Introduction and Related Works

Background

Proposed Method

PCAg

Simulation Conclusion

DPCA

Proposed Method PCAg Simulation Conclusion DPCA Distributed Principal Component Analysis Assumptions: Each

Distributed Principal Component Analysis

Assumptions:

Each sensor receives sensed value from its neighbors Only the data of sensors which are in the radio range of each other are correlated

For the computation of eigenvectors, DPCA uses:

v

t+1

i

=

j=1 C ij v j

N

t

Each sensor broadcasts its eigenvector’s entry v i and its sensed value.

Assumptions of DCPA are not supported by weather data

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 12
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
12 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

Using PCA for Data Reduction

PCA Reconstruction Error Using PCA for Data Reduction DML Retained Variance for 50 sensors Retained Variance

DML

Retained Variance for 50 sensors Retained Variance λ 1 /λ 2 Temperature 1.0 3.6909e+015 Humidity
Retained Variance for 50 sensors
Retained Variance
λ 1 /λ 2
Temperature
1.0
3.6909e+015
Humidity
1.0
3.4407e+015
Light
1.0
2.4402e+015
Voltage
1.0
3.1737e+015
Combined
1.0
1.9305e+015
.
.
.
.
.
.
.
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
13 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Leaf nodes send their observations

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Intermediate nodes compute PCs . .BS ˆ.x = gy
Intermediate nodes compute PCs . .BS ˆ.x = gy .y = f (x 1 ,
Intermediate nodes
compute PCs
.
.BS
ˆ.x = gy
.y = f (x 1 , x 2 , x 3 , x 4 , x 5 , x 6 , x 7 , x 8 )
Intermediate nodes
recalculate PCs according
to reconstruction error
.n 8
.x 6
.y 1 = f (x 1 , x 2 , x 5 )
.y 2 = f (x 3 , x 4 , x 7 )
The next level
intermediate nodes
reconstruct observation
and compute PCs from all
downstream observations
.n 6
.n 5
.n 7
.x 1
.x 2
.x 3
.x 4
.n 1
.n 2
.n 3
.n 4
The base station
reconstructs observations
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
14 / 34
.

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Updating Eigenvectors

Eigenvectors are common between each pair of parent-child intermediate nodes

Nodes which are in the vicinity of the base station must send large eigenvectors.

For these nodes, the computation of the covariance matrix and eigenvectors is energy consuming

Updated eigenvectors must be sent to the base station

The base station calculates final eigenvectors

.

DML

8 Dec 2010

Amirmohammad Rooshenas

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error .BS .g . 8 . g = g 8
.BS .g . 8
.BS
.g
.
8

.g = g 8

.
.

.

.n 8 . .g 5 .g 7 . . .n 5 .n 7 .n 6
.n 8
. .g 5
.g 7
.
.
.n 5
.n 7
.n 6
.
.
.
.
.n 1
.n 2
.n 3
.n 4
. . . . . . Reducing The Data Trans. in WSN using PCA 15
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
15 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Features of Using PCA

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Features of Using PCA Each intermediate node has error
PCA LocalPCA: Using Collective PCA Reconstruction Error Features of Using PCA Each intermediate node has error

Features of Using PCA Each intermediate node has error threshold that can be configured by the base station

High threshold value causes low accuracy and high efficiency

An intermediate node only sends PCs in each epoch

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 16
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
16 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Collective PCA

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error DML Reconstruction of data in each intermediate node and

DML

Reconstruction of data in each intermediate node and combination again is not scalable

Collective PCA: PCA is invariant to an orthonormal linear transformation

We want to find the principal component of x = [x where s is sensed value.

y 1 and y 2 are principal compenents of x 1 and x 2

T

1

, x

T

2

, s] T

z =

y 1

y 2

s

  and g is the eigenvector of zz T then

the principal component of x is y = g T z

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 17
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
17 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Aggregation of PCs using Collective PCA Leaf nodes send

Aggregation of PCs using Collective PCA

Leaf nodes send their observations

ˆx. = gy Intermediate nodes compute PCs . .BS .y = f (y 1 ,
ˆx.
= gy
Intermediate nodes
compute PCs
.
.BS
.y
= f (y 1 , y 2 , x 6 , x 8 )
The next level
intermediate nodes
compute PCs from
observations and received
PCs
.n 8
.y 1 = f (x 1 , x 2 , x 5 )
.x 6
.y 2 = f (x 3 , x 4 , x 7 )
.n 5
.n 6
.n 7
The base btation
reconstructs observations
.x 1
.x 2
.x 3
.x 4
.n 1
.n 2
.n 3
.n 4
Intermediate nodes only
send PCs in each epoch
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
18 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Updating Eigenvectors

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Each intermediate node has its eigenvectors T T .g
Each intermediate node has its eigenvectors T T .g = [g 8 (1) ∗ g
Each intermediate node
has its eigenvectors
T
T
.g
= [g 8 (1) ∗ g
, g 8 (2) ∗ g
, g 8 (3), g 8 (4)] T
5
7
.BS
.
Each intermediate node
updates its eigenvectors
independently
.g 5 , g 7 , g 8
.
.n 8
Updated eigenvectors must
be sent to the base station
.
.
.g 5
.g 7
.
The base station calculates
final eigenvectors
.n 6
.n 5
.n 7
.
.
.
.
Update rate depends on
the number of changes in
environment
.n 1
.n 2
.n 3
.n 4
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
19 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Proposed method: An aggregation service for calculation of

Proposed method: An aggregation service for calculation of reconstruction Error

In PCAg, the base station needs to update eigenvectors in order to compute reconstruction error

In LocalPCA, the base station needs to tune thresholds

In PCAg and LocalPCA, the base station do not access the real observation to calculate reconstruction error

We proposed an aggregation service facilitating the calculation of reconstruction error at the base station.

EAPCAg : We enhance PCAg using this aggregation service

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 20
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
20 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Using PCA LocalPCA: Using Collective PCA Reconstruction Error

PCA LocalPCA: Using Collective PCA Reconstruction Error Proposed method: An aggregation Service for Calculation of

Proposed method: An aggregation Service for Calculation of Reconstruction Error

ˆx x 2 =

x x) T x x)

= ˆx T ˆx ˆx T x x T ˆx + x T x

=

x T AA T A A T x 2x T AA T x + x T x

T A A T x − 2 x T AA T x + x T x
T A A T x − 2 x T AA T x + x T x

I

= x T x − x T AA T x = x T x −
= x T x − x T AA T x
= x T x − y T y
=
n p
2
y
2 p ≪ n
x i
i
i=1
i=1
.
.
.
.
.
.

DML

8 Dec 2010

Amirmohammad Rooshenas

Reducing The Data Trans. in WSN using PCA

21 / 34

Data Set

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results 7500 epoch of 10 sensors from Intel-lab . DML 8

7500 epoch of 10 sensors from Intel-lab

.
.

DML

8 Dec 2010

Amirmohammad Rooshenas

.

.

.

.

.

Reducing The Data Trans. in WSN using PCA

.
.

22 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Simulation Senario

Settings

Results

Simulation Conclusion Simulation Senario Settings Results DML 8 Dec 2010 Amirmohammad Rooshenas . . . .
Simulation Conclusion Simulation Senario Settings Results DML 8 Dec 2010 Amirmohammad Rooshenas . . . .

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 23
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
23 / 34

Packet

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results Packet Size   Field Size Header 10

Packet Size

 

Field

Size

Header

10

Observation

2

Eigenvector

4

element

Principal

4

Component

Squared

4

observation

Request for

1

.

observation

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 24
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
24 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Numerical Results

Settings

Results

Simulation Conclusion Numerical Results Settings Results Numerical Results Transmitted Error Error Error

Numerical Results

Numerical Results Transmitted Error Error Error Parameters Bytes Mean Variance Max PCAg 200340 0.82 0.96 8.34
Transmitted Error Error Error Parameters Bytes Mean Variance Max PCAg 200340 0.82 0.96 8.34 update:
Transmitted
Error
Error
Error
Parameters
Bytes
Mean
Variance
Max
PCAg
200340
0.82
0.96
8.34
update: 11 epoch
LocalPCA
168114
0.82
0.07
1.43
threshold at 23:1.0
threshold at 22,21: 0.35
EAPCAg
215216
0.82
0.23
6.16
threshold:1.42

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 25
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
25 / 34

Discussion

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results For the sensor node 23rd, EAPCAg method has sent about

For the sensor node 23rd, EAPCAg method has sent about 15K bytes more than PCAg algorithm

PCAg method has sent about 32K bytes more than LocalPCA.

The average updating interval in EAPCAg is about 13.84 epochs which is around 3 epochs more than PCAg.

EAPCAg has less efficiency than PCAg because the size of each PC packet is increased from 14 bytes to 18 bytes

In PCAg and EAPCAg methods, when the number of sensor nodes increases, the eigenvector packets should be divided to smaller packets, so the amount of transmitted data will increase

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 26
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
26 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results Comparision of LocalPCA with PCAg at the Base Station .

Comparision of LocalPCA with PCAg at the Base Station

. . . . . . . DML 8 Dec 2010 Amirmohammad Rooshenas Reducing The
.
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
27 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results Error of LocalPCA . . . . . . .

Error of LocalPCA

. . . . . . . DML 8 Dec 2010 Amirmohammad Rooshenas Reducing The
.
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
28 / 34

Outline Introduction and Related Works Proposed Method Simulation Conclusion

Settings

Results

Proposed Method Simulation Conclusion Settings Results . Error of PCAg . Error of EAPCAg . .

.

Error of PCAg

.

Error of EAPCAg

.

.

.

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 29
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
29 / 34

Outline Introduction and Related Works

Future Works

Proposed Method

References

Simulation Conclusion

Appendix

Future Works

Simulation Conclusion Appendix Future Works PCA Efficiency of LocalPCA depends on thresholds

PCA

Efficiency of LocalPCA depends on thresholds

Thresholds should be adjusted adaptively

Weather data has low dimensions. LocalPCA will show better performance with high dimensional data (e.g. image )

Kernel PCA is an non-linear version of PCA and can handle data which changes rapidly

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 30
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
30 / 34

Outline Introduction and Related Works

Future Works

Proposed Method

References

Simulation Conclusion

Appendix

Method References Simulation Conclusion Appendix Borgne, Y. L., Raybaud, S., and Bontempi, G. Distributed

Borgne, Y. L., Raybaud, S., and Bontempi, G. Distributed principal component analysis for wireless sensor networks. Sensors 8 (2008) Hillol, K., Huang, W., Krishnamoorthy, S., Byung-Hoon, P., and Wang, S. Collective principal component analysis from distributed, heterogeneous data. Lecture Notes in Computer Science. 2000.

Jollife, I. Principal Component Analysis, second ed. Springer,

2002.

Y. L. Borgne and G. Bontempi. Unsupervised and supervised compression with principal component analysis in wireless sensor networks. in Proc. First International Workshop on Knowledge Discovery from Sensor Data, 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2007

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 31
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
31 / 34

Outline Introduction and Related Works

Future Works

Proposed Method

References

Simulation Conclusion

Appendix

Method References Simulation Conclusion Appendix Questions? . . . . . . . DML 8 Dec

Questions?

. . . . . . . DML 8 Dec 2010 Amirmohammad Rooshenas Reducing The
.
.
.
.
.
.
.
DML
8 Dec 2010
Amirmohammad Rooshenas
Reducing The Data Trans. in WSN using PCA
32 / 34

Outline Introduction and Related Works

Future Works

Proposed Method

References

Simulation Conclusion

Appendix

How to find eigenvectors

Simulation Conclusion Appendix How to find eigenvectors DML Computation of Covariance Matrix Σ = E [(

DML

Computation of Covariance Matrix Σ = E [(x E [x])(x E [x]) T ]

Solve eigenvector, eigenvalue equation Σα = λα

if λ i is largest eigenvalue and Σα i = λ i α i then α i is dominant eigenvector.

When the dimension reduction from n dimension to q dimension is efficient?

Retained Variance: H(q) = q

k=1 λ k

n

k=1 λ k

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 33
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
33 / 34

Outline Introduction and Related Works

Future Works

Proposed Method

References

Simulation Conclusion

Appendix

Power Iteration Method

Simulation Conclusion Appendix Power Iteration Method Calculation of eigenvectors by eigenvector, eigenvalue

Calculation of eigenvectors by eigenvector, eigenvalue equations is time consuming O(n 3 ).

Power Iteration Method approaches to dominant eigenvector

very fast O(n 2 ).

PIM iteration: g k+1 = Σg k

FastPCA : Extension of PIM for calculation of other eigenvectors.

DML

8 Dec 2010

Amirmohammad Rooshenas

. . . . . . Reducing The Data Trans. in WSN using PCA 34
.
.
.
.
.
.
Reducing The Data Trans. in WSN using PCA
34 / 34