h a l f b a k e r yOn the one hand, true. On the other hand, bollocks.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
It's a common idea for autonomous vehicles to
communicate constantly with a server to best
function in traffic. If a server knows that 600 cars
want to go from the same area A to the same area B it
can tell all of them which roads to use, allocating more
cars to faster and wider roads. This
way all the cars
reach their destinations as quickly and safely as
possible. This real world traffic routing is highly
analogous to information routing in networks.
A common objection to having a car that
communicates constantly with a server is the
potential for abuse and the ability to constantly track
each person's car. It violates one's right to privacy.
I propose a method for cars to take advantage of city-
wide traffic coordination without owners suffering
loss of privacy. This protocol would flow as follows:
An autonomous vehicle is turned on and given a
destination. At this point it generates a random IP6
network address. The car connects to the city's traffic
network and sends its approximate size, desired
maximum and minimum average speed to destination,
desired top speed, and destination.
The city's traffic server uses traffic cameras and usage
history to decide how many of the vehicles on the
street are taking directions. Allocating for more or less
variance in uncontrolled traffic it sends packets
addressed to the car with directions. As the car
travels it sends information on its location and desired
course. Accidents and traffic jams can be avoided as
well by communicating with randomly generated ID's
as static ones.
Being open source the protocol can be used by a wide
variety of software. By allowing any car to ask for
directions at any time it can provide service regardless
of underlying architecture.
It's robust against many forms of DOS attack because
the server can tell when cars are actually present with
actual traffic flows. If a car lies about its destination
the most it can do is make the server plan an
unnecessary route. If there are more requested routes
than available roads (due to a known DOS attack) the
server can still route cars as efficiently by over-
allocating roads. It won't matter which cars are real
and which are part of the DOS attack. If there are
really too many cars for the roads allocation degrades
gracefully.
[link]
|
|
//autonomous vehicles to communicate constantly with a server |
|
|
"Blimey, bit parky innit" |
|
|
"Yeah, tell me about, all the diodes down my left hand side are aching, but continue on to GPS xxxxxxxx, who've you got on board?" |
|
|
"Some nerk from Eastenders, trying to make it to the studio. Gahh, humans eh? Tell you what, your make a traffic jam and let's have a bit of fun" |
|
|
The concept is good, but I'm not sure if it should be entirely open source access. Some level of identification is desirable to pinpoint witnesses, or culprits, in traffic accidents. Likewise, knowing who it is that drives the wrong way up a one-way street, speeds, or runs a red light. That data is valuable and shouldn't be lightly given up. |
|
|
I think it's a title awaiting an idea. Too long. Geez
almost a paragraph just reading the name. |
|
| |