Top Banner
Rescue Base Station Ibrahim Ghaznavi, Kurtis Heimerl, Umar Muneer , Abdullah Hamid, Kashif Ali, Tapan Parikh, Umar Saif A drop-in communication platform for disaster struck regions
40
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Rescue Base Station

Rescue Base Station

Ibrahim Ghaznavi, Kurtis Heimerl, Umar Muneer , Abdullah Hamid, Kashif Ali, Tapan Parikh, Umar Saif

A drop-in communication platform for disaster struck regions

Page 2: Rescue Base Station

Special Thanks!

OpenBTS & Freeswitch community Endaga, Range Networks, Fairwaves

Google Faculty Research Awards National Science Foundation- NSF

Page 3: Rescue Base Station

Disasters can happen: Any time

Anywhere Any magnitude

Page 4: Rescue Base Station

Kashmir Earthquake Affected 3.5 Million

Page 5: Rescue Base Station

Indian Ocean Tsunami affected 1.7 Million people across 22 countries

Page 6: Rescue Base Station

Typical Consequences

Page 7: Rescue Base Station

Restoration can take weeks

Page 8: Rescue Base Station

Problem:

• Disruption of Information and Communication.• Hundreds of thousands of victims in trouble.• No way to effectively channelize the relief efforts• Communication needs:

– Victims with victims– Rescue workers with rescue workers– Rescue workers with victims– Relatives of victims with victims

• Race against time to save lives

Page 9: Rescue Base Station

Network Requirements - Base

• Quickly Deployable• Uses Existing Phones• Off-Grid• Robust to node failures• Autonomous distributed operation

– Can’t expect people to create point-to-point links

Page 10: Rescue Base Station

Network Requirements - Services

• Broadcast Announcements• Connect people

– Victim/victim, worker/worker– Outside/victim, victim/worker– Intelligently, as user availability may vary

• Identify skills/resources in population– E.g., doctors and blood type– Searchable

Page 11: Rescue Base Station

The Big idea

• Portable, solar-powered cellular base stations– Call, SMS and GPRS

• Different nodes establish a disconnected mesh– Cellular network meet Sneakernet– Users as mules to exchange data

Page 12: Rescue Base Station

The Rescue Base Station

Page 13: Rescue Base Station

Inside an RBS

• OpenBTS• Freeswitch• Gateway GPRS Support Node – GGSN• Serving GPRS support Node –SGSN• Packet Control Unit - PCU

Page 14: Rescue Base Station

How it works

Page 15: Rescue Base Station

Single RBS Unit

Coverage Range

RBS

Page 16: Rescue Base Station

User Registration

• Any user can register with our network bysending the following information as an SMSto 111:– Contact Number– Name– Occupation– Blood Group

Page 17: Rescue Base Station

Example

Page 18: Rescue Base Station

Eventual Consistency

• A mechanism to synchronize user registrationand call record data across the disconnectedthe RBS(s)

Page 19: Rescue Base Station

Multiple RBS units working together

Page 20: Rescue Base Station

Why Eventual Consistency? •To gather information from the neighboring nodes and use it connect via intelligent call routing (greater the information greater are the chances of connectivity).

•To boost the search capability via Attribute based search (more records synced, better search results).

•So that the user do not have to re-register at other neighboring RBS sites

Page 21: Rescue Base Station

Data Packet Creation

A packet is created if: •The total number of user registrations in a BTSis greater than 500

•Or the total registration records are less than500 and an time-interval of 30 mins has lapsed.

Page 22: Rescue Base Station

• The packet deletion is only possible if thenumber of nodes in the RBS system are fixed

• Each chuck has 2 data variables:–Hop count–Node count

• If the hop count equals to the node count itmeans that the packet has been synced acrossall the BTSes and it can be deleted.

Data Packet Deletion

Page 23: Rescue Base Station

Out-of-the-box Services

• Intelligent call routing• Attribute based search• Emergency voice-mail• Emergency shortcodes• Emergency Alerts• Dynamic addition/ removal of RBS(es)

Page 24: Rescue Base Station

Intelligent call routing

RBS

Step 2: RBS called User 2 who is not

available

Step 1: user1 makes a call to

user2

Step 3: RBS automatically calls the ‘most recent caller’ who called

user2

Page 25: Rescue Base Station

Attribute Based Search

• An SMS based yellow page like look-up service on the following attributes: – Name – Occupation – Blood group

Page 26: Rescue Base Station

Examples: Attribute based search

Page 27: Rescue Base Station

Emergency Voice-mail

• This enable a user to record his specificcomplaint directly to the rescue team (inhis/her vicinity)

• By listening to these complaints rescue teamscan be channelized in an organized fashion.

Page 28: Rescue Base Station

Emergency Shortcodes

• You may directly call the following professionals at the following shortcodes: – Doctors (7777 ) – Police (7700 ) – Fire brigade (7722 ) – Rescue Workers (911 )

Page 29: Rescue Base Station

Short-code Logic Example: Doctors (7777 )

Doctor 1

Doctor 2

Doctor 3

Doctor 4

Doctor 5

Calling Doctor 1; Failed attempt

Calling Doctor 2; Failed attempt

Calling Doctor 3; Failed attempt

Calling Doctor 4; Successful attempt

Page 30: Rescue Base Station

Other Features

• Emergency Alerts: An emergency announcement can be broadcasted to all registered users ( in- range of the local RBS) via an SMS.

• Dynamic addition/ removal: We can add and remove RBS nodes to and from the system to efficiently meet the variations in connectivity demand of a particular disaster struck site.

Page 31: Rescue Base Station
Page 32: Rescue Base Station

Evaluation

• An In-Lab Experiment• Simulations

– Nodes VS synchronizations– Packet size VS synchronizations– People VS synchronizations

Page 33: Rescue Base Station

Lab Evaluation

Page 34: Rescue Base Station

Lab Evaluation

Page 35: Rescue Base Station

Simulation

• All the simulation are tested on the followingtwo network topologies:

• Mesh topology• Line topology

Page 36: Rescue Base Station

Nodes VS Synchronizations

Page 37: Rescue Base Station

Packet size VS Synchronizations

Page 38: Rescue Base Station

People VS Synchronizations

Page 39: Rescue Base Station

Future Work

• We are collaborating with a US basedcompany 'Endaga' to release acommercial version of the Rescue BaseStation.

• We plan to (but hope not to) test it on a largerscale in a real-world setting.