Kommunikationssysteme FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE FGAN 0 Relay Placement for Ad-hoc Networks in Crisis.

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

1

Relay Placement forAd-hoc Networks in Crisis and

Emergency Scenarios

RTO IST-086 Symposium on C3I for Crisis, Emergency and Consequence Management

Bucharest, Romania

11. May 2009

Thorsten Aurisch, Jens Tölle

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

2Overview

• Challenge, Basic Scenario• Ad-hoc Network Properties• Indoor Communication Challenges• Relay Placement Concepts• Demonstrator• Test Scenarios• Results• Improvements• Conclusions

• Challenge, Basic Scenario• Ad-hoc Network Properties• Indoor Communication Challenges• Relay Placement Concepts• Demonstrator• Test Scenarios• Results• Improvements• Conclusions

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

3The Challenge – Basic Scenario

Emergency Responder/

Unites

Crisis/Emergency Situation

Incident Command Center/

Command Post

Command and Control

Feedback, Sensor Data,

ACKs, Blueforce Tracking,…Crisis Area

Reliable two-way communication needed!

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

4

crisis area

mobilebase station

units

incident commandcenter

crisis area

ad-hocnetwork

incident commandcenter

units

Communication Structures

Base stations:

GSM, UMTS, PMR

- Single Hop

- Infrastr. dependant

Mobile Ad-hoc

networks

- Multi Hop

- not Infrastr. dependant

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

5Ad-hoc Network Advantages

No infrastructure needed

No installation of temporary base station needed

Units can start communicating directly upon arrival

Additional units enhance density of the ad-hoc network:- multiple routes- increased robustness

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

6

commandpost

interruption

Ad-hoc Network Challenges

Sometimes difficult radio propagation

properties, especially in buildings

Risk of communication network

partitioning

Rapid deployment needed in previously

unknown terrain

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

7

interruption

t1

t2

Indoor Challenge

signal strength

strong weak

Sudden decrease of signal strength (corners, concrete walls, metal doors) may cause network disruption

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

8

relay node

Indoor Challenge: Solution Relay Placement

Signal strength is monitored,

decreasing signal strength triggers alarm,

unit activates and deploys relay node.

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

9Relay Placement: Known Concepts

Relay placement approaches are used in different areas:

Sensor networks:

Usage of relay nodes to improve connectivity

Energy efficiency by decreasing transmission power

Autonomous multi-robots:

Robots can actively move to positions to (re)gain connectivity

Tracking by means of pre-installed relay nodes:

Limited range of application

Sometimes used: Optimization algorithms with unrealistic assumptions

Needed: Pragmatic and robust solution

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

10The RelayP Concept

Assumptions

• Relay nodes are simple, cheap, and available

• Usage of COTS components (e.g. IEEE 802.11 b/g radios)

• Incident environment and current situation unknown

• Exisiting communication infrastructure failed or overloaded

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

11The RelayP Concept II

Demonstrator development:

- Decision to drop a node depends on quality of existing link- Used metric: Received Signal Strength (RSS)- Two different methods for RSS calculation

1. Determine RSS by radiotap header analysis of last N beacons

- Beacons are periodically sent by all devices

- Approach does not generate additional traffic

2. Usage of channel probing signals- Send PROBE_MSG every seconds- Deployed relay node or incident command centre replies with

PROBE_ACK- Calculate moving average of last N messages

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

12Test Scenario 1/3

Demonstrator test cases:

(Notebooks, Linux 2.6.26, IEEE802.11b)

Outcome of tests depends on building, materials,…

Test scenario 1:

Communication in a straight line

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

13Test Scenario 2/3

Test scenario 2:

Communication moving

around a corner

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

14Test Scenario 3/3

Test scenario 3:

Communication entering a room

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

15Signal Strength vs. Packet Loss

Estimation of threshold

-96 dbm

radio reception threshold

-90 dbm as

relay drop

threshold?

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

16Attenuation by moving persons

Estimation of Deviation

Result:

Standard deviation of 1.5 dbm

-87 dbm relay drop threshold

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

17Distance Measurements

Distance [m]

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

18Drawbacks and Potential Improvements

• Organize common movement of response units

• Consider small-scale fluctuations when moving in groups by combined evaluation of metric

• Consider network partitioning in case of group movements

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

19Conclusions

• Relay placement approach for crisis scenarios

• Fast and pragmatic solution

• No optimal solution, but sufficiently good heuristic

• No preparation time needed

• Does not constrain emergency responders/units

Kommunikationssysteme

FORSCHUNGSINSTITUT FÜR KOMMUNIKATION, INFORMATIONSVERARBEITUNG UND ERGONOMIE

FGAN

20…

Thank you very much for your Attention! Questions?

Jens Tölle

toelle@fgan.de

Telephone:

+49 228 9435 513

Fax:

+49 228 9435 685