Energy Conservation in

waralligatorMobile - Wireless

Nov 21, 2013 (3 years and 4 months ago)

74 views

Energy Conservation in
wireless sensor networks


Kshitij Desai,
Mayuresh

Randive
,

Animesh

Nandanwar







Basic Design


Sensor Network Architecture






Internet

Sensor
Network

Sink

Architecture of a Sensor Node













Ref: Energy Conservation in Wireless Sensor Networks


a
Survey

Observations


Communication Sub
-
system consumes more energy than
computation sub
-
system


Energy to transmit one bit = Energy for execution 1000
. instructions


Radio component requires same order of energy for
reception, transmission and idle states


Sensing sub
-
system might also require significant amount
of energy based on the type of sensor node.

Three Main enabling Techniques


Duty
-
cycling


Data
-
Driven approaches


Mobility

Duty
-
cycling



Topology Control


Power Management


Sleep/Wake Protocols


On
-
demand, scheduled rendezvous and
Async


MAC Protocols with low Duty
-
cycle


TDMA, Contention
-
based and hybrid

Data
-
driven approaches


Data reduction


In
-
Network Processing


Data
-
Compression


Data
-
prediction


Stochastic, Time
-
series Forecasting and algorithmic
approaches


Energy
-
efficient data acquisition


Adaptive Sampling


Hierarchical Sampling


Model
-
Driven active sampling


Mobility
-
based

approaches


Mobile
-
sink


Mobile
-
relay

ATPC: Adaptive Transmission

Power Control for Wireless

Sensor Networks

Main Points


What is this paper about?


Power saving for wireless communication


Paper style?


Empirical study + a little theory work


What is the contribution?


Study of spatial
-
temporal impact on communication


Mechanism to adaptively achieve an optimal transmission power
consumption


Motivation

11

TP1

TP2

TP2

Motivation

12

TP1

TP1

T1

T2

The minimum transmission power level

to save energy

and
maintain specified link quality

TP2

T2

Design Goals


Achieve energy efficiency


The minimum transmission power



Maintain Link Quality


Reliable links



In runtime systems, dynamic environments


Spatial impact


Temporal impact

13

Roadmap

Data Analysis

Empirical Observation

Algorithm Design

Algorithm Evaluation

PART 1

PART 2

Part 1
-
Transmission Power vs. Link Quality


Link Quality Metrics


RSSI (Received Signal Strength Indication), LQI (Link
Quality Indication), and PRR (Packet Reception Ratio)


Transmission Power Level Index (3~31)


Experiments on Spatial Impact


5 pairs of motes, 3 environments


100 packets at each transmission power level


RSSI/LQI/PRR measured at different distances

15

Part 1
-

Investigation of Spatial Impact





-95
-90
-85
-80
-75
-70
-65
-60
-55
-50
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Transmission Power Level Index
RSSI (dbm)
2 ft
6 ft
12 ft
18 ft
24 ft
28 ft
-95
-90
-85
-80
-75
-70
-65
-60
-55
-50
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Transmission Power Level Index
RSSI (dbm)
3 ft
6 ft
12 ft
18 ft
24 ft
30 ft
16

(a) RSSI measured on a grass field

-95
-90
-85
-80
-75
-70
-65
-60
-55
-50
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Transmission Power Level Index
RSSI (dbm)
3 ft
6 ft
12 ft
18 ft
24 ft
30 ft
(c) RSSI measured in a parking lot

1.
Different shapes at the same distance in
different environments

2.
Different degree of variation in different
environments

3.
Approximately linear

(b) RSSI measured in a corridor

Investigation of Temporal Impact


Experiment on Temporal Impact


In brushwood where human activity is rare, over 72 hours


9 MicaZ motes in a line, 3 feet apart


A group of 20 packets at each power level every hour

-95
-93
-91
-89
-87
-85
-83
-81
-79
-77
-75
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Transmission Power Level Index
RSSI (dbm)
9am 1st Day
10am 1st Day
11am 1st Day
12pm 1st Day
1pm 1st Day
2pm 1st Day
17

-95
-93
-91
-89
-87
-85
-83
-81
-79
-77
-75
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Transmission Power Level Index
RSSI (dbm)
0am 1st Day
8am 1st Day
4pm 1st Day
0am 2nd Day
8am 2nd Day
4pm 2nd Day
1. Vary gradually but noticeably over time

2. Approximately parallel

(a) RSSI measured every 8
-
hour

(b) RSSI measured every hour

Part 1
-

Link Quality Threshold





18

Binary link quality thresholds

Slight different in different environments

(a) RSSI Threshold on a grass field

(b) LQI Threshold on a grass
field

0
20
40
60
80
100
120
-95
-90
-85
-80
-75
-70
RSSI (dbm)
PRR (%)
0
20
40
60
80
100
120
50
60
70
80
90
100
110
LQI (Reading from MicaZ)
PRR (%)
Part 2
-

Model Design of ATPC


Use a linear model to approximate a non
-
linear correlation


rssi(tp) = a ∙ tp +b


Least
-
square


approximation


Dynamic model


a and b vary


from time to time

19

Part 2
-

ATPC
Overview





0.4TP+32
6
4
0.8TP+49
27
3
0.5TP+23
12
2
Control Model
Power Level
NodeID
0.4TP+32
6
4
0.8TP+49
27
3
0.5TP+23
12
2
Control Model
Power Level
NodeID
ATPC Table at Node
1
20

Initialization Phase: build models from linear approximation

Node
3
Node
5
Node
4
Node
1
Node
2
Transmission Range
Runtime Tuning Phase: pairwise closed loop control

Packet with
Transmission Power
Level
12
Notification
25

8

Part 2


Closed Loop Control

Start
here

RSSI, LQI
and PRR

Part 2
-

Experiment Setup

22


Current transmission power control algorithms


A node
-
level non
-
uniform solution (Non
-
uniform)


Network
-
level uniform solutions

»
Max transmission power level (Max)

»
The minimum transmission power level over nodes in a network
that allows them to reach their neighbors (Uniform)


A 72
-
hour continuous experiment with MicaZ


A spanning tree of 43 nodes, 24 leaf nodes


Leaf nodes send 32 packets to the base every hour

Part 2
-

Experimental
Setup

23

(a) Weather Conditions over 72 Hours

(b) Spanning Tree Topology

(c) Experimental Site

Part 2
-

Packet Reception Ratio

0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0
6
12
18
24
30
36
42
48
54
60
66
72
Time (hours)
End-to-end PRR
ATPC
Max
Uniform
Non-Uniform
0
10
20
30
40
50
60
70
80
90
100
0
6
12
18
24
30
36
42
48
54
60
66
72
Time (hours)
PRR (%)
Link with Static
Transmission
Power
Link with ATPC
24

(a)
E2E packet reception ratio


Max ~ 100%



ATPC ~ 98.3%



Uniform ~ 98.3%



Non
-
Uniform ~ 58.8%


(b) PRR at a chosen link

ATPC ~ constantly 100%

Static transmission power


~ vary from 0% to 100%

Part 2
-

Transmission Energy Consumption

0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
6
12
18
24
30
36
42
48
54
60
66
72
Time (hours)
Relative Transmission Energy
Consumption
ATPC
Max
Uniform
Non-Uniform
25

Max ~ 100%


ATPC ~ 58.3% (1% control overhead)

Uniform ~ 68.6%


Non
-
Uniform ~ 43.2%


Relative energy consumption

Conclusions and Future Work


Benefits of ATPC lie in three core aspects:


ATPC maintains above 98% E2E PRR over time


ATPC achieves significant energy savings


53.6% of the transmission energy of Max


78.8% of the transmission energy of Uniform


ATPC accurately adjusts the transmission power


Adapting to spatial and temporal factors



Towards reliable and energy
-
efficient
routing


Spatial reuse for concurrent transmissions

26






Questions?









27

Thank you very much!