An Analysis of Energy Usage of Energy Efficient Protocols for Routing for Mobile Ad Hoc Networks

1
An Analysis of Energy Usage of Energy Efficient
Protocols for Routing for Mobile Ad Hoc Networks

• Presented by
• Jay Elston
• November, 2007

2
Contents

• Project Goals Accomplishments
• New and Improved
• Analysis Results
• Issues Encountered
• Conclusions
• Questions

3
Project Goals(What I wanted to do)

• Duplicate Analysis Results for EAGER
• Improve on EAGER if possible
• Compare EAGER with improved version of EAGER and
  other Hybrid approaches (such as ZRP)
• Use Network Simulation to validate the analysis

4
Accomplishments (What I actually did)

• (Almost) Duplicated Analysis Results for EAGER
• Two improvements to EAGER
• EAGER-C a modification to the geometry of the
  cell periphery
• EAGER-LR use Location management to make
  inter-cell messaging more efficient
• Compared EAGER, EAGER-C and EAGER-LR

5
Loose Ends

• My Analysis did not get the same values that
  EAGER got
• Probably culprit the Radio Model
• Did not compare EAGER to ZRP
• Ran out of time ?.
• Did not get to the simulation phase
• Ran out of time ?.

6
Key Accomplishments

• Analysis of EAGER
• Including effects of radio model and node density
  changes.
• Two EAGER Improvements

7
EAGER Improvements
EAGER uses a trapezoidal peripheral area. Minimum
Rc is the cell edge.
EAGER-C uses a semi-circle as the peripheral
area. When Ap is small, so is Rc.
8
EAGER Improvements
EAGER floods the network with location requests
to find a target node.
EAGER-LR uses a location registry to find out
which cell the target node is in.
9
Analysis Results

• Ran with parameters as close a possible to EAGER
  paper
• EAGER paper compared EAGER with Reactive and
  Proactive approaches
• I added comparisons with EAGER-C and EAGER-LR
• I also adjusted some simulation parameters

10
Comparison of results
My EAGER levels out too soon ? But the overall
shape is similar ?
11
Results
N 30000
R 1000
Etx Ercv 50E-10 Etx Ercv 50E-10
Ed 10E-12 Ed 10E-12
N 30000
R 1000
Etx Ercv 50E-9 Etx Ercv 50E-9
Ed 10E-10 Ed 10E-10
Changing the radio power by 10x changes the
numbers, but the overall relationships remain ?
12
Results
N 60000
R 1000
Etx Ercv 50E-10 Etx Ercv 50E-10
Ed 10E-12 Ed 10E-12
N 7500
R 500
Etx Ercv 50E-10 Etx Ercv 50E-10
Ed 10E-12 Ed 10E-12
Changing the node density or the size of the
network also changes the numbers, but the overall
relationships remain ?
13
Results
EAGER-C offers a marginal improvement to EAGER
EAGER-LR also offers improved performance until
EAGER goes to a single cell.
14
Issues Encountered

• Duplicating Results
• Not all factors are documented
• For the EAGER analysis
• What were the parameters used for the radio?
• How did they analyze the reactive and proactive
  routing protocols?
• Time
• Was not able to do all analysis I wanted to do
• Did not get a chance to learn how to use a
  network simulator

15
Future Work

• Add analysis of ZRP
• If EAGER is much better than ZRP, then look at
  how to improve EAGER to fix problems with
  node-free areas
• Network simulation
• More analysis of use of location registries
• Note this is a motivating scenario for
  migrating services ?

16
Conclusions

• Location registries can offer some energy
  efficiency improvements for MANETs.
• The big issue with energy efficient routing in
  MANETs is establishing routing tables this
  dwarfs the other factors.

17
References

• Zhao, Qing Tong, Lang Counsil, David
  Energy-Aware Adaptive Routing for Large-Scale Ad
  Hoc Networks Protocol and Performance Analysis,
  Mobile Computing, IEEE Transactions on Volume 6, 
  Issue 9,  Sept. 2007 Page(s)1048 1059

18
Questions?