Energy Efficiency in

gilamonsterbirdsΗλεκτρονική - Συσκευές

24 Νοε 2013 (πριν από 3 χρόνια και 4 μήνες)

53 εμφανίσεις

Energy Efficiency in
Ad
-
Hoc Networks

Chien
-
Jen Huang

& Min Chen

April 22, 2004


EE497A

2

Outline


Introduction


Physical layer design


Routing layer design


Cross
-
layer design issues


Conclusions

3

Introduction


What is Ad
-
hoc networks


Energy efficiency is one of the crucial
problems for Ad
-
hoc



Energy constraint strategy is a global
issue, involving

most
layer stacks in the
network structure


4

Physical layer design

--
System model


The

whole

transmission period time



The total energy consumption for transmitting bits







The energy
-
constrained modulation problem finally can be modeled
as
:


Minimize




Subject to
delay constraint



peak
-
power constraint

tr
sp
on
T
T
T
T



E
L
tr
tr
sp
sp
on
on
T
P
T
P
T
P
E



tr
tr
on
c
t
T
P
T
P
P




)
)
1
((

L
E
E
bit
/

tr
on
T
T
T



max
0
max
)
1
(
0
P
P
P
c
t





tr
tr
sp
sp
on
c
amp
t
T
P
T
P
T
P
P
P





)
(
5

Physical layer design


--
Energy constraint problem for M
-
QAM



An upper bound of the symbol error probability for M
-
QAM in AWGN
channel is given in [5] as






SNR



Free space propagation model



We obtain the lower bound of the transmission energy as


2
1
2
3
)
2
1
1
(
4
)
1
2
3
(
)
2
1
1
(
4









b
e
Q
Pe
b
b
b
f
r
N
B
P
2
2



r
r
t
t
P
G
G
L
d
P
2
2
2
)
4
(



on
d
b
BT
L
f
on
t
t
BT
G
P
N
T
P
E
on
BT
L
on
)
2
1
(
4
ln
)
1
2
(
3
4
2
2






6

Physical layer design


--
Energy constraint problem for M
-
QAM
(cont’d)


The total energy consumption per bit as





Peak
-
power constraint


we can find a lower bound such that all satisfy the
peak
-
power constraint




Energy consumption optimal problem for M
-
QAM is modeled as



Minimize


Subject to



The optimal modulation constellation size is obtained from


by

L
T
P
T
P
T
P
E
tr
tr
on
c
on
t
total
/
)
)
1
((





c
t
P
P
P



max
)
1
(

min
T
min
T
T
on

total
E
T
T
T
T
T
tr
on




max
min
*
on
BT
L
b

b
M
2

*
on
T
7


Physical layer design


--
Numerical results











Fig. 1 versus Fig. 2 versus

T
T
on
/
total
E
total
E
b
8

Physical layer design

--
Further discussions



M
-
QAM versus M
-
FSK



QAM sacrifices transmit power to obtain higher bandwidth
efficiency


FSK sacrifices bandwidth for a reduction in transmit power




Coded versus uncoded modulation


Error control coding can reduce the required SNR for the same
performance requirement


Possible bandwidth expansion caused by error control coding
redundancy, and the extra energy consumption of the encoder


Power Control And
Routing Protocol In
Network Layer

10

Design Rules For Power Control
Routing Protocol


Routing with more
hops can save power


Many short hop may be
better than a few long ones


Routing with more
hops reduces
interference at every
node

hop
BC
AB
BC
AB
hop
Pt
Td
Td
d
d
T
Pt
Td
Pt
CGtGr
T
Td
Pt
d
CGtGr
Pt
2
1
Pr
Pr
)
(
Pr
2
,
Pr
1
1
Pr



















Power P required to transmit over a distance d

11

Common Power Routing
Protocol (COMPOW)



Characteristics


Equal transmission power between every
node in the route


The common power is the minimal one which
can keep the connectivity of the route


Compatible to some other protocol



Address resolution protocol (ARP)



DHCP



Reverse ARP

12


COMPOW will be not efficient when the node in
the network is not uniformly distributed



13

Cluster Power Routing Protocol
(CLUSTERPOW)



Characteristic


Use power level clustering the nodes into
groups


The cluster may last for several levels


Every node in the same cluster level transmits
the same power to each other


Designed based on “distance vector routing”
algorithm

14

Property Of CLUSTERPOW


Clusterpow is adaptive to the network
distribution


The route is discovered with a non
-
decreasing
power levels


COMPOW is the special case of Clusterpow


Clusterpow at every fixed power level is loop
free


Clusterpow is loop free


15

Clusters in the ad hoc network

[7] V. Kawadia and P. R. Kumar, Power Control and Clustering in
Ad Hoc Networks,
IEEE INFOCOM 2003
.

16

Power Efficiency Issue In Cross
Layer



Device power issue



Increase hops number also increase the
device power consumption ( contradict to the
design rule)


Error propagation issue




simply relay case



Data is just amplified without any demodulation or
decoding when passing by the intermediate nodes
between source and destination

17

18

Power Efficiency Issue In Cross
Layer


regenerator case



Data will be demodulated, decoded or some other
processing which can improve the error rate when
passing by the intermediate nodes between source
and destination


Trade off


Use regenerators may decrease the error rate but
also increase the processing power at every node


With error propagation, more hops are not
necessarily good


19

Simulation1:

Compare the cases of directly transmission, simply relay and
regenerator

4
-
QAM, pass loss exponent=4, dac=5m ,y=0m

2 hops routing, equal error rate in the final nodes

20

Simulation2:

Compare the cases of directly transmission, simply relay and
regenerator

4
-
QAM, pass loss exponent=4, dac=5m ,y=2m

2 hops routing, equal error rate in the final nodes

21

Conclusion


Optimizing the performance of each layer
is not necessarily optimizing the global
performance


Achieving the global performance, cross
layer issues should be taken into account



22



Questions?



23

References


[1] A. J. Goldsmith and S. B. Wicker, Design Challenges for Energy
-
constrained Ad Hoc Wireless
Networks,
IEEE Wireless Communications Magazine
, pp. 8
-
27 Aug. 2002.


[2] A. Y. Wang, S. Chao, C. G. Sodini, and A. P. Chandrakasan, Energy Efficient Modulation and
MAC for Asymmetric RF Microsensor System,”
International Symposium on Low Power
Electronics and Design
, pp. 106
-
111, 2001.


[3] S. Cui, A. J. Goldsmith, and A. Bahai, Energy
-
constrained Modulation Optimization under
Energy Constraints,

ICC’03
, Alaska, U.S.A, May, 2003.


[4] S. Cui, A. J. Goldsmith, and A. Bahai, Energy
-
constrained Modulation, submitted to IEEE
Trans. on Wireless Communications.


[5] J. G. Proakis,
Digital Communications,

4th Ed. New York: MxGraw
-
Hill, 2000.


[6] S. Hares, H. Yanikomeroglu, and B. Hashem, A Relaying Algorithm for Multihop TDMA TDD
Networks Using Diversity,
IEEE Vehicular Technology Conference 2003 (
VTC'F03
)
, Orlando,
Florida, USA, October. 2003.


[7] V. Kawadia and P. R. Kumar, Power Control and Clustering in Ad Hoc Networks,
IEEE
INFOCOM 2003
.


[8] Vikas Kawadia and P. R. Kumar, A Cautionary Perspective on Cross Layer Design, submitted
to
IEEE Wireless Communication Magazine
. July, 2003.


[9] Z. Sahinoglu and P. Orlik, Regenerator Versus Simple
-
Relay With Optimum Transmit Power
Control for Error Propagation,
IEEE Communications Letters
,

Vol. 7,

Issue 9,

pp. 416
-
418,
September 2003.


[10] Z. Sahinoglu,and P. Orlik, Optimum Power Compensation for Error Propagation in Relay
Assisted Wireless Networks,
IEEE Global Telecommunications Conference
, Vol. 1, pp. 382
-
386,
December 2003.