The SkyNet project

The original idea came to me when I saw an old magazine photograph that showed a pyramid building with an intense laser beam coming out of its top to the sky. Then the beam was split on to the sky and formed a laser grid. This was a science fiction picture but it was the first time I thought about a network that would use open space laser beams to transfer data. Optical fibers are used to transfer huge amounts of data but they are still cables. The idea of data over a laser beam in the air is not new. There are already commercial solutions up to several Mbps and also good amateur solutions like the Ronja project for 10Mbps. Laser data links have some very important advantages over wire and radio communications, so they must be considered as great solutions for specific applications:

Although laser links seem to be the solution to many of our problems, they have some disadvantages that is worth to be mentioned:

From the currently available amateur solutions, Ronja seems to be the best among them. The only problem with Ronja is that it is too complicated for the amateur to build and it costs too much. It is a link of high data rate and long distance though. In my approach for the SkyNet I thought that it might be more affordable and easier for many applications, to have more than two smaller and cheaper transceiver modules to cover the same distance (repeaters) rather than a big one. On the other hand, if talking about backbones Ronja may be better, but if talking about a crowded neighborhood where we want to interconnect the houses, then we may not need so advanced electronics, neither so precise alignment.

There is also another thing that must be considered when working with laser data links. All the commerce and amateur links that I have seen are point to point links. There are just two transceivers that communicate with each other. No third transceiver can join the link as the laser beam is highly directive. I did not mention it as a disadvantage of the laser links because I intend to propose a few solutions to this problem.

Solution 1: The passive way
Using special lenses, called beam splitters, a single laser beam can be split into two other. In an ideal beam splitter each produced beam is half the power of the original beam. Beam splitters can provide a way of making a laser link network, nevertheless they may be considered as impractical solution for the next reasons:

Solution 2: The active way
A more practical networked approach for laser links seems to be the use of active repeaters. Each repeater accepts a laser beam and transmits a copy (or more than one copies) of the original data. By the use of active repeaters many of the problems of passive repeaters can be overcame.

Each repeater may be a back-to-back transceiver or a more simple and cost effective hardware solution that just repeats the received data, without further data processing.

I have talked so far about repeaters that connect a pair of transceivers, but how can multi node networks be created? Most of the home and company networks today use the star topology. Star topology is good because the network is up, even if one or some of the nodes are offline. It allows all nodes to communicate with each other through a central point (called hub in computer terminology). The critical point now is the hub and not the nodes. A user has to only take care about its link with the hub and not the links to the other users. If we need to connect two star networks together then we can use a backbone between the two hubs, or we can connect them using another central node. The last case introduces the tree network and it is used for greater scale networks.

For SkyNet, the star network topology seems to be satisfactory since we want the network to be always available, regardless of the availability of the nodes and since the links are relatively of small distance. Later on, if we want to connect two sub networks, we can make a higher data rate and reliability backbone between their two hubs. In order to implement this topology, there must be a central hub with as many transceivers as the clients. Each transceiver in the hub must point to a specific client.

(to be continued)

Current files for download:
Laser transceiver schematic
(Express PCB file)
Laser transceiver schematic (jpg file)


Disclaimer notice: Information and files in this page provided as is. You are free to copy any images and data in this page  provided that you do not alter their content and you clearly define their source.
Copyright 2009-2010 - All Rights reserved