dizzysoul
June 14th, 2011, 11:44 PM
Hello. I'm looking to develop a new technology for field deploying many internet nodes in an area and having them quickly interconnect to each other. I'm not looking to take a mesh networking approach, but rather something different.
This project is going to develop a cheap and effective open-source solution for building and deploying a basic network infrastructure to provide internet access and basic networking tools in an emergency situation. The project will use as many cheap and on-the-shelf components as possible, simple assembly, and I'm looking to have the designs and software for the finished project released to the public for anyone to use. (like the RepRap project)
I've seen other projects developed with the same goal, and I have studied them, but I think I can offer a much better solution.
I made this thread because I wanted to start a discussion on this project.
I want people to ask me questions!
I want people to poke holes and point out flaws!
I want people to offer suggestions, or even link articles to related projects!
I'm looking to get some expert opinions and advice on my idea. Critics and cynics are welcome here! I want to discuss what's practical, impractical, what works/doesn't work, what parts can be thrown out, elaborated on, and what can be improved. Please let me know what you think!
I'm 25 and I've been doing computer, low voltage and IT work for over 10 years. I've been interested in developing this system for humanitarian efforts because of disasters like Haiti and internet blackout situations like Egypt and Libya. Any and all advice is welcome, thank you!
Smart Path Logic Internet Node for Tactical and Emergency Response (SPLINTER)
Overview:
SPLINTER is designed to be a fast, cheap, mass deploy-able internet-in-a-box. Each unit is equipped with dual-band WiFi, GPS, mobile broadband, dual-core computer, rechargeable battery and solar cell. Depending on the configuration, you can add disk storage, a satellite uplink, and more.
Each unit is designed to provide routing and internet access to existing infrastructure, or to nearby personal devices using its powerful built-in WiFi antennas. Each unit acts as a network node, using a second wireless channel to communicate and route traffic to other nearby units. With location-aware “Smart Path Logic” technology, each unit can build a geographical map of other units and use this information to trace efficient network routes across many different nodes, without using less effective brute-force methods and mesh networking protocols.
Each unit also monitors the status of itself and other nearby nodes. Using a “push-pull” system, nodes can quickly relay changes to each other and across the network. This means the network is self-healing; being able to repair routes and adapt to constant changes in the network topography.
Each unit is equipped with a dual-core computer and a suite of network, internet and hosting software. This includes DNS, LAMP, POP/IMAP, SMB, CMS, and more. These services can be configured to activate dynamically based on the overall “health” of the network. If too many DNS servers go down, for example, other nodes can step in to pick up the workload, even if they are configured to do something else. Each unit comes with a limited amount of flash storage, with the option to add high capacity disk storage for web hosting and other media.
The unit itself is rugged, shock proof and weather proof. It’s lightweight, and can be deployed via a number of methods, including airdrop. It contains a number of connection ports, including power, ethernet, and external antenna connectors. The box also contains basic indicators for battery level, signal strength, and network status. When not connected to external power, the box uses solar panels (and optional hand crank) to charge the internal battery. The unit can also carry additional extension poles and cables anchors (to be assembled on-site) allowing the unit to extend the antenna platform an additional 18 feet.
SPLINTER uses a simple, modular design to cut down on costs. Each unit is highly configurable to the needs of many different scenarios, from disaster relief, to nation building, to military applications. All software is Linux based; It can be audited and configured to meet specific security standards.
How it works:
Let's say you load 1,000 units into a squadron of aircraft, and you blanket a geographical area with these units (with a tight enough spread so they can interconnect, or some ground coordination to move units into optimal position) in order to build the initial network.
The units will behave as follows:
"Where am I?" - use it's GPS, and possibly other sensors, to find it's exact geographical location.
"Where are my friends?" - will broadcast it's location, and listen for other broadcasts, either using wifi or a more basic radio technology.
"Have you met any new friends?" - will consolidate a list of discovered units, and rebroadcast this list to nearby nodes. When a new unit is discovered, goes missing, or changes location, the list is updated and rebroadcast.
This basic logic will propagate very quickly, like a large game of marco-polo, until every units knows the geographical location of every other unit. This will create a realistic 'topography' of the entire network, realized by each and every device. When a change in the topography is detected, that change is "pushed" throughout the network using the above logic. Even if a "push" update is missed, units will periodically check-in with other nearby units, and "pull" any changes that result.
Each node will have a unique subnet, and when combined with the topography map, this means network traffic will have a sense of direction. It's like having a navigation system in each packet of data; It knows where to go, it knows which paths to take, and even if it makes a wrong turn, it knows how to recalculate the route. The overall network can also be optimized using a 'hierarchy' algorithm. Based on flow of traffic, signal strength, signal overlap, and other factors, certain nodes can be ranked higher than others. This will create multi-tiered node clusers, and even network "superhighways" to handle larger volumes of traffic.
In other words, the network will work to organize itself into a system that mirrors what we use today, the world wide web.
Potential issues:
WiFi isn’t strong enough to properly interconnect between units. Either the power will have to be dramatically increased, possibly beyond FCC specifications and power availability through solar energy. Another solution is to add directional and (possibly) motorized antennas to the antenna cluster. Directional antennas can have a range of many miles under low power conditions. This means the units might require some elevation, optional pole extensions, and some level of on-the-ground assembly.
The solar panels might be inefficient, or require expensive and exotic materials to meet the power demands of the router equipment and computer. One solution is to limit transmission power, network bandwidth and processing power to conserve battery, but this would make it more difficult to interconnect nodes and host media (such as video and audio).
Other features:
The optional hand crank will use a large gear ratio to maximize the charging of the dynamo. The hand crank will be robust and detachable so that any kind of mechanism can spin the crankshaft.
All of the components inside of each unit are recyclable and can be salvaged or re purposed for other tasks. The computer and networking equipment can be flashed and programmed to serve other functions. The solar panels, dynamo, and battery will not be wired in-line with the power circuit, and can be removed as long as the unit is plugged into an external power source, much like how you can remove a laptop battery after it's plugged in.
This project is going to develop a cheap and effective open-source solution for building and deploying a basic network infrastructure to provide internet access and basic networking tools in an emergency situation. The project will use as many cheap and on-the-shelf components as possible, simple assembly, and I'm looking to have the designs and software for the finished project released to the public for anyone to use. (like the RepRap project)
I've seen other projects developed with the same goal, and I have studied them, but I think I can offer a much better solution.
I made this thread because I wanted to start a discussion on this project.
I want people to ask me questions!
I want people to poke holes and point out flaws!
I want people to offer suggestions, or even link articles to related projects!
I'm looking to get some expert opinions and advice on my idea. Critics and cynics are welcome here! I want to discuss what's practical, impractical, what works/doesn't work, what parts can be thrown out, elaborated on, and what can be improved. Please let me know what you think!
I'm 25 and I've been doing computer, low voltage and IT work for over 10 years. I've been interested in developing this system for humanitarian efforts because of disasters like Haiti and internet blackout situations like Egypt and Libya. Any and all advice is welcome, thank you!
Smart Path Logic Internet Node for Tactical and Emergency Response (SPLINTER)
Overview:
SPLINTER is designed to be a fast, cheap, mass deploy-able internet-in-a-box. Each unit is equipped with dual-band WiFi, GPS, mobile broadband, dual-core computer, rechargeable battery and solar cell. Depending on the configuration, you can add disk storage, a satellite uplink, and more.
Each unit is designed to provide routing and internet access to existing infrastructure, or to nearby personal devices using its powerful built-in WiFi antennas. Each unit acts as a network node, using a second wireless channel to communicate and route traffic to other nearby units. With location-aware “Smart Path Logic” technology, each unit can build a geographical map of other units and use this information to trace efficient network routes across many different nodes, without using less effective brute-force methods and mesh networking protocols.
Each unit also monitors the status of itself and other nearby nodes. Using a “push-pull” system, nodes can quickly relay changes to each other and across the network. This means the network is self-healing; being able to repair routes and adapt to constant changes in the network topography.
Each unit is equipped with a dual-core computer and a suite of network, internet and hosting software. This includes DNS, LAMP, POP/IMAP, SMB, CMS, and more. These services can be configured to activate dynamically based on the overall “health” of the network. If too many DNS servers go down, for example, other nodes can step in to pick up the workload, even if they are configured to do something else. Each unit comes with a limited amount of flash storage, with the option to add high capacity disk storage for web hosting and other media.
The unit itself is rugged, shock proof and weather proof. It’s lightweight, and can be deployed via a number of methods, including airdrop. It contains a number of connection ports, including power, ethernet, and external antenna connectors. The box also contains basic indicators for battery level, signal strength, and network status. When not connected to external power, the box uses solar panels (and optional hand crank) to charge the internal battery. The unit can also carry additional extension poles and cables anchors (to be assembled on-site) allowing the unit to extend the antenna platform an additional 18 feet.
SPLINTER uses a simple, modular design to cut down on costs. Each unit is highly configurable to the needs of many different scenarios, from disaster relief, to nation building, to military applications. All software is Linux based; It can be audited and configured to meet specific security standards.
How it works:
Let's say you load 1,000 units into a squadron of aircraft, and you blanket a geographical area with these units (with a tight enough spread so they can interconnect, or some ground coordination to move units into optimal position) in order to build the initial network.
The units will behave as follows:
"Where am I?" - use it's GPS, and possibly other sensors, to find it's exact geographical location.
"Where are my friends?" - will broadcast it's location, and listen for other broadcasts, either using wifi or a more basic radio technology.
"Have you met any new friends?" - will consolidate a list of discovered units, and rebroadcast this list to nearby nodes. When a new unit is discovered, goes missing, or changes location, the list is updated and rebroadcast.
This basic logic will propagate very quickly, like a large game of marco-polo, until every units knows the geographical location of every other unit. This will create a realistic 'topography' of the entire network, realized by each and every device. When a change in the topography is detected, that change is "pushed" throughout the network using the above logic. Even if a "push" update is missed, units will periodically check-in with other nearby units, and "pull" any changes that result.
Each node will have a unique subnet, and when combined with the topography map, this means network traffic will have a sense of direction. It's like having a navigation system in each packet of data; It knows where to go, it knows which paths to take, and even if it makes a wrong turn, it knows how to recalculate the route. The overall network can also be optimized using a 'hierarchy' algorithm. Based on flow of traffic, signal strength, signal overlap, and other factors, certain nodes can be ranked higher than others. This will create multi-tiered node clusers, and even network "superhighways" to handle larger volumes of traffic.
In other words, the network will work to organize itself into a system that mirrors what we use today, the world wide web.
Potential issues:
WiFi isn’t strong enough to properly interconnect between units. Either the power will have to be dramatically increased, possibly beyond FCC specifications and power availability through solar energy. Another solution is to add directional and (possibly) motorized antennas to the antenna cluster. Directional antennas can have a range of many miles under low power conditions. This means the units might require some elevation, optional pole extensions, and some level of on-the-ground assembly.
The solar panels might be inefficient, or require expensive and exotic materials to meet the power demands of the router equipment and computer. One solution is to limit transmission power, network bandwidth and processing power to conserve battery, but this would make it more difficult to interconnect nodes and host media (such as video and audio).
Other features:
The optional hand crank will use a large gear ratio to maximize the charging of the dynamo. The hand crank will be robust and detachable so that any kind of mechanism can spin the crankshaft.
All of the components inside of each unit are recyclable and can be salvaged or re purposed for other tasks. The computer and networking equipment can be flashed and programmed to serve other functions. The solar panels, dynamo, and battery will not be wired in-line with the power circuit, and can be removed as long as the unit is plugged into an external power source, much like how you can remove a laptop battery after it's plugged in.