Actions

Mesh Networking: Difference between revisions

From HacDC Wiki

No edit summary
 
(9 intermediate revisions by 4 users not shown)
Line 1: Line 1:
Meetings Thursdays at 7:30ish starting 10 Feb 2011
Meetings Thursdays at 7:30ish starting 10 Feb 2011
=Brainstorming=
=Brainstorming=
==Official Project Page==
[[Byzantium | Byzantium]]
The notes below were used while formulating the project. Further notes available on the now-Ongoing project page.
==Directions==
==Directions==
Which way will we head with this?
Which way will we head with this?
Line 9: Line 15:
*Communication technology which lends itself to improvisational construction.
*Communication technology which lends itself to improvisational construction.


=What Next?=
==What Next?==
After we have our direction what are we gonna do?
After we have our direction what are we gonna do?
*define coverage
*define coverage
Line 22: Line 28:
*purpose - what will this be used for (surfing the web or just email/twitter/<insert favorite text service here>)
*purpose - what will this be used for (surfing the web or just email/twitter/<insert favorite text service here>)
*Use case - circumstances under which this technology will usually be deployed
*Use case - circumstances under which this technology will usually be deployed
**ISP/ISP replacement
**disaster
**disaster
**communication infrastructure failure
**communication infrastructure failure
**censorship
**censorship


=Aproaches=
==Aproaches==
==Mesh==
===Mesh===
What will our approach to the network level of the mesh be?
What will our approach to the network level of the mesh be?
*broadcast
*broadcast
*"actor" oriented - assume no knowledge in other nodes
*"actor" oriented - assume no knowledge in other nodes
*nested - roll with what exists and use it to connect seperate traditional nets  
*nested - roll with what exists and use it to connect seperate traditional nets
*Hybrid
**clients communicate using wifi with hubs
**hubs relay traffic using one of the mesh networking protocols
**IPv6 makes the addressing for this very sensible. address: [1st half highest node "subnet"][further refinement or subnets][last block device id unique within the subnet]


==Extension==
===Extension===
If we need to augment our mesh to cover our target area how will we do that?
If we need to augment our mesh to cover our target area how will we do that?
*"same stuff but on a larger physical scale"
*"same stuff but on a larger physical scale"
*a higher traditional network
**mesh nodes with more broadcast power
**some mesh nodes also have point-to-point links with each other
***tight-beam RF
***optical
****laser
****incoherent light
*a higher order traditional network


=Crazy Ideas=
==Crazy Ideas==
*viral node generation - present captive portals on carriers that require the user to become a "infected" and themselves present captive portals to other non infected devices (zombie smartphone apocalypse)
*viral node generation - present captive portals on carriers that require the user to become a "infected" and themselves present captive portals to other non infected devices (zombie smartphone apocalypse)
*make it easy and fun for people to join
**mesh networking app in appstores
**add something fun to make them want to install the app, like a game or a visualization
**as long as they play once in a while, they route traffic (tamagochi effect)
*one-stop shop router reflashing
**flash your Linksys WAP, automatically have mesh networking support active in the background
*TonidoPlug?
*datacomm over
**shortwave
***Note: requires more than two hams to participate
**FRS (been done before but would be an interesting experiment)
**GMRS (been done before but would be an interesting experiment)
**laser pointer with a photocell receiver
**incoherent light [http://ronja.twibright.com/ Ronja]
**Sneakernet with rsync-like functionality (project: mnemonic?)
**Slideshow
**One relatively large video screen
**One relatively decent webcam (not hard)
**Video screen displays sequences of 2D barcodes
**Computer with webcam reads and decodes barcodes
***Reminds me a bit of Conway's Game of Life as a network technology.
*Fidomail-like transfer and synchronization of data bundles
**dialup
**tight-beam datacomm of some kind
**sneakernet (project: mnemonic?)
**Cross-technology implementation
***Web forums
***Microblogs
***Blogs
***BBS software
*Automated sending and decoding of Morse code
**Might not be all that fast but would be easy to implement and fun to test the bandwidth of.
*How could we abuse SMS for this?
*Mobile nodes
**Someone sets up a (small) node in their car and drives around slowly.
**Set up in a backpack or under a coat
***On bicycle
***On rollerblades
***On skateboard
***ON UAV
***On a balloon


=Project=
=Project=
Line 48: Line 106:
==Journal==
==Journal==
[[MeshNetworkingMeeting_20110210]]
[[MeshNetworkingMeeting_20110210]]
{{Template:Byzantium}}

Latest revision as of 21:16, 3 April 2012

Meetings Thursdays at 7:30ish starting 10 Feb 2011

Brainstorming

Official Project Page

Byzantium

The notes below were used while formulating the project. Further notes available on the now-Ongoing project page.

Directions

Which way will we head with this?

  • support Networking_Strategic_Simulation
  • community wifi
  • cold hard maxout of the current tech
  • Development of communication devices with readily available (i.e., cheap) technology.
  • Communication technology which lends itself to improvisational construction.

What Next?

After we have our direction what are we gonna do?

  • define coverage
    • building
    • block
    • neighborhood
    • city
    • intercity
    • regional
    • infinite (assuming it will grow with the technology's ability to support it)
  • audience - who's gonna use this
  • purpose - what will this be used for (surfing the web or just email/twitter/<insert favorite text service here>)
  • Use case - circumstances under which this technology will usually be deployed
    • ISP/ISP replacement
    • disaster
    • communication infrastructure failure
    • censorship

Aproaches

Mesh

What will our approach to the network level of the mesh be?

  • broadcast
  • "actor" oriented - assume no knowledge in other nodes
  • nested - roll with what exists and use it to connect seperate traditional nets
  • Hybrid
    • clients communicate using wifi with hubs
    • hubs relay traffic using one of the mesh networking protocols
    • IPv6 makes the addressing for this very sensible. address: [1st half highest node "subnet"][further refinement or subnets][last block device id unique within the subnet]

Extension

If we need to augment our mesh to cover our target area how will we do that?

  • "same stuff but on a larger physical scale"
    • mesh nodes with more broadcast power
    • some mesh nodes also have point-to-point links with each other
      • tight-beam RF
      • optical
        • laser
        • incoherent light
  • a higher order traditional network

Crazy Ideas

  • viral node generation - present captive portals on carriers that require the user to become a "infected" and themselves present captive portals to other non infected devices (zombie smartphone apocalypse)
  • make it easy and fun for people to join
    • mesh networking app in appstores
    • add something fun to make them want to install the app, like a game or a visualization
    • as long as they play once in a while, they route traffic (tamagochi effect)
  • one-stop shop router reflashing
    • flash your Linksys WAP, automatically have mesh networking support active in the background
  • TonidoPlug?
  • datacomm over
    • shortwave
      • Note: requires more than two hams to participate
    • FRS (been done before but would be an interesting experiment)
    • GMRS (been done before but would be an interesting experiment)
    • laser pointer with a photocell receiver
    • incoherent light Ronja
    • Sneakernet with rsync-like functionality (project: mnemonic?)
    • Slideshow
    • One relatively large video screen
    • One relatively decent webcam (not hard)
    • Video screen displays sequences of 2D barcodes
    • Computer with webcam reads and decodes barcodes
      • Reminds me a bit of Conway's Game of Life as a network technology.
  • Fidomail-like transfer and synchronization of data bundles
    • dialup
    • tight-beam datacomm of some kind
    • sneakernet (project: mnemonic?)
    • Cross-technology implementation
      • Web forums
      • Microblogs
      • Blogs
      • BBS software
  • Automated sending and decoding of Morse code
    • Might not be all that fast but would be easy to implement and fun to test the bandwidth of.
  • How could we abuse SMS for this?
  • Mobile nodes
    • Someone sets up a (small) node in their car and drives around slowly.
    • Set up in a backpack or under a coat
      • On bicycle
      • On rollerblades
      • On skateboard
      • ON UAV
      • On a balloon

Project

Goal

Spec

Journal

MeshNetworkingMeeting_20110210

Project Byzantium
Main Meeting Notes Byzantium Live Distro Misc. Development Stuff