I spend odd amount of time periodically thinking about ways in which I could send information. This time I’d like to loosen the constraints and say that I’m ok with sending information without strong guarantees of delivery. I’d like send this data cheaply over a dynamically generated network (“ I don’t know the destinations before hand or who/what is in the network”) . The network should be mobile, and without gateways or hubs that organize all the nodes. The network wouldn’t be very fast or necessarily reliable in terms of transmitting to a final destination. However, it might be more fault tolerant given you had alot of these data receivers and generators in a close vicinity. I think what I’m attempting to define is a Lazy Network. A network that’s not necessarily optimized for speed and has limited guarantees on delivery, where the cost of entering or leaving a network is marginal. So the following is my attempt to explore this idea of a Lazy Network a little more.
When I think about Lazy Networks, I try and imagine why they might be a good idea. I think about why I’d need or want something to be lazy.
Cookies are what’d I’d consider to operate within a lazy network. The cost of having a cookie or not is low, the delivery guarantee of recieving additional information from that cookie is not guaranteed, and the speed in which you recieve additional information from that cookie is not necessarily fast. The flow for cookies is as follows.
Here we have bees that visit various flowers, and lazily distribute the pollen to the location in which they feed or fly around, there are no guarantees that the pollen reaches a habitable environment, but the “message” is carried none the less.
I think that the idea of a Lazy network might make sense in terms of doing low bandwidth out of 802.11 communication. If you wanted to build a network that allowed for a probabilistic distribution of messages. The deliverability factor could be dictated by the number of nodes nearest the distribution point. For example if you wanted to run a deal or live coupon as a store owner. You might consider using an out of band message hub to broadcast your deal message. Using this softer network , you might limit your distribution to a relevant radius. Additionally, if you wanted to make a cookie for the physical world, you might broadcast this near field or out of band signal, bluetooth even and rely on the ad hoc nodes, to simply append or add to the cookie, until it found its way near a hub or a gateway or returned to the point of origin. Allowing you to gather an accurate history of the device. One other application might be an emergency broadcast system, that used lower frequency out of band communication, that would be cheap and highly redundant in high device areas. Think major metropolitan areas. ### Where am I headed ? I think that more thought should be given to using out of band messages, to create and maintain real world networks with high mobility, that are accurate by means of the physical link that is afforded to the device. A device outside of 50ft can’t be apart of the network, simply because the link probabilistically wouldn’t be able to send that message. You might find that you’re able to connect people or send relvant information without gps.