12 September 2006 1

Utilization of Commercial Wireless Networking

Technology in Simulated Martian Environments

Stephen Horan, Phillip DeLeon, Deva Borah, VishwanathChukkula, Vijay Velusamy, Gaylon Lovelace, Anirudh Daga,

Sumaharsha Yelamanchi, Ray LymanKlipsch School of Electrical and Computer Engineering

New Mexico State University

This material is based upon work supported by the National Aeronautics and Space Administration under Award No. NAG3-2864.

12 September 2006 2

Topics

• Project Motivation & Goals• Modeling the RF Environment on Mars• Protocol Simulation Suite• Performance Evaluation of the 802.11a

Standard• Performance Evaluation of the 802.11b

Standard• Conclusions & Recommendations

12 September 2006 3

Project Motivation & Goals

• NASA plans for future planetary exploration include the use of wireless networks in sensor web and planetary rover applications

• Research questions:– What are the limitations (power, range, data rate) of

commercial wireless network technologies (designed for indoor use) in such a planetary application?

– Is the outdoor multipath so strong as to prevent useful operation?

– Is there an advantage to using IEEE 802.11a vs. 802.11b? If so, is a general advantage or only under specific terrain conditions?

– What is the network performance as a function of the number of nodes?

12 September 2006 4

Project Motivation & Goals

• Simulate the RF environment of proposed landing sites on Mars using digital elevation models (DEMs) from Mars Global Surveyor and RF planning/ propagation software tools; use current primary landing sites at Gusev Crater and Meridiani Planumas test subject areas for study and modeling

• Understand limitations such as power, range, data rate, BER of COTS wireless networking technology when utilized on planetary surface

• Propose modifications to COTS wireless technology that would enable reliable networking of nodes on a planetary surface

12 September 2006 5

• ATDI’s HerTZMapper and ICS Telecom– Software for

radio network planning, spectrum management, and radio coverage analysis

– RF models take into account environmental data andsystem parameters (code parameters modified for Mars)

Modeling the RF Environment on Mars

12 September 2006 6

Modeling the RF Environment on Mars

• RF coverage patterns for Gusev1, Site 1– Green denotes -84dBm– Red denotes -93dBm

• Site Coverage = 32.42% (without clutter)19.55% (with clutter)

• Maximum Coverage Distance, dmax =1137m (without clutter)1185m (with clutter)

12 September 2006 7

Modeling the RF Environment on Mars

• In addition to the power studies, the simulation software was used to generate Power Delay Profiles (PDF) for various locations.

• Simulations were validated against measurements made locally.

0 2 4 6 8 10 12 14 16 18 20

-300-260-220-180-140-100

Hematite-5 Site 1 RMS Delay Spread = 0.15 μs;Received Power = 42 nW;

0 2 4 6 8 10 12 14 16 18 20

-300-260-220-180-140-100

Time (μs)

Hematite-4 Site 1 RMS Delay Spread = 0.07 μs;Received Power = 23 nW;

0 2 4 6 8 10 12 14 16 18 20

-300-260-220-180-140-100

Time (μs)

Pow

er a

t Rec

eive

r (d

Bm

per

10

ns)

Gusev-1 Site 3 RMS Delay Spread = 7.09 μs;Received Power = 1.3e-005 nW;

0 2 4 6 8 10 12 14 16 18 20

-300-260-220-180-140-100

Time (μs)

Gusev-1 Site 2 RMS Delay Spread = 0.47 μs;Received Power = 0.011 nW;

0 2 4 6 8 10 12 14 16 18 20

-300-260-220-180-140-100

Time (μs)

Gusev-1 Site 1 RMS Delay Spread = 0.07 μs;Received Power = 62 nW;

Response continues beyond 45 μs...

12 September 2006 8

Protocol Simulation Suite

• The PDP’s generated form the basis for the IEEE 802.11 a/b simulation studies to determine the performance of the physical and medium access layers in this environment.

• Used the following tools for performance simulations:– mWLAN toolbox for MATLAB to simulate the physical layer

performance for the IEEE 802.x protocols– OPNET run as a co-simulation for the networking

performance simulation; calls mWLAN for physical layer performance

12 September 2006 9

MAC Layer Simulation Methodology

Physical Layer

simulation in MATLAB

MAC Layer simulation in

OPNETStatistics

DelayLoadThroughputRetransmission Attempts

Source and Destination

Address

Received power

Power Delay

Profiles

ICS Telecom

Digital Elevation MapsGusev CraterMeridiani Planum (Hematite)

ITM Propagation

Model

OPNET-MATLAB Co-Simulation

12 September 2006 10

Performance Evaluation of the 802.11a Standard

BER for Gusev 1, Site 1

0 10 20 30 40 50 600

5

10

15

20

25

30

Simulated Time (mins)

MA

C D

elay

( μse

c)

Gusev1 Site 1Gusev1 Site 2Hematite4 Site 1Hematite5 Site 1

0 10 20 30 40 50 60

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Simulated Time (mins)Th

roug

hput

per

uni

t Loa

d

Gusev1 Site 1Gusev1 Site 2Hematite4 Site 1Hematite5 Site 1

Network MAC delay and relative throughput at different sites.

12 September 2006 11

Performance Evaluation of the 802.11a Standard

• Networking Results:– Pretty Good Overall– Gusev1 Site 1 is

Mediocre• Opposite of PHY Trend

– At Hematite4:• 3,4,5 Nodes – All Okay• Tx Pwr ≥ 100 µW – All

Okay• 100 vs 1024 Bytes/Pkt –

Both Okay• More Traffic Delay

Only

• Physical Layer Results:– Error Rates are “Not

Bad” for Dist < 500 m– PER < 0.1 is Easy for

MAC to Handle• Multipath Dominates

Noise, for PTX > 1 mW– Antenna Height Helps,

for Very Low Power• Might Hurt, for Higher

Power– Behavior Very Location

Specific• Hills & Valleys

12 September 2006 12

Performance Evaluation of the 802.11b Standard

BER for Gusev 1, Site 1

0 10 20 30 40 50 600

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

Simulation time (min)

Del

ay (s

ec)

Gusev1-Site1Gusev1-Site2Hematite4-Site1Hematite5-Site1

Network MAC delay at different sites.

12 September 2006 13

Performance Evaluation of the 802.11a Standard

• Physical Layer Results:– Increase in antenna height does

not improve performance significantly for 802.11b within 0.5 m to 2 m.

– RAKE receiver improves performance significantly for 802.11b.

• Networking Results:– Large MAC layer delay due to

significant number of retransmissions. Throughput per unit load is severely limited by multipath.

– Increasing number of nodes from 3 to 5 does not significantly affect delay and slightly decreases throughput per unit load for low packet arrival rates.

– Large packet sizes increase delay and decrease throughput per unit load. However, there is less energy per successful bit.

– Increase in power from 1 mW to 1 W does not improve MAC layer performance significantly.

12 September 2006 14

Conclusions & Recommendations

• The use of commercial link planning software can be successfully used to model rover-type performance on the surface; validation is necessary to make sure that the parameters are set properly.

• The simulations showed two definite regions: a power-limited region and a multipath-limited region.– As expected, increasing power in the power-limited region

helps until the multipath-limited region is reached.• Simulations showed a great deal of variation from site

to site as the local topology changed. Location needs to be accounted for in route planning.

12 September 2006 15

Conclusions & Recommendations

• IEEE 802.11a has good Physical Layer performance up to a few hundred meters; lower data rate shows better BER than high data rate

• IEEE 802.11b was more sensitive to the multipath effects– a RAKE receiver improved performance

• IEEE 802.11a had better MAC performance than 802.11b– Packet size, data rates, retry rates, and other

parameters can be selected to tweak results.

12 September 2006 16

Conclusions & Recommendations

• Recommendations– Re-clock IEEE 802.11a for long multipath

management (extends guard interval at the cost of halving the data rate)

– Investigate IEEE 802.11g (lower carrier frequency alone can give improvement)

– Look into coming standards such as IEEE 802.11n and IEEE 802.16 that are being designed for mutipath and/or outdoor links.