> If the access points are spaced closely enough, and if one can hand you
> off to another quickly enough, and if they are distant enough from the
> roadway, then you can probably drive at highway speeds without being
> disconnected. But those are an awful lot of ifs.

Here's one way to think about this. Assume a coverage overlap of C, and a
car velocity of V.

If pre-authentication + reassociation can take place in time T = C/V then
you can have seamless handoff without disconnection.

In other words Vmax = C/T, where Vmax is the maximum velocity at which
seamless handoff can be achieved.

Let's assume for a moment, that we have open authentication, so that
security doesn't enter into the calculations. From Bill Arbaugh's recent
measurements, the reassociation time alone may range between 40 and 400
ms.

With a coverage overlap C = 5 feet, and T = 40 ms, then

Vmax = 125 ft/second = 85 mph. However, with T = 400 ms, then
Vmax = 12.5 ft/second = 8.5 mph.

So a lot depends on the reassociation times.

Note that where Vmax > D/T where D = total coverage Diameter, then you
will be out of the coverage area entirely by the time reassociation is
completed. So at velocities greater than Vmax 802.11 can no
longer be used. Typically this velocity is characteristic of plane flight
(0.1 - 0.3 Mach) rather than vehicular travel, though.


--
general wireless list, a bawug thing <http://www.bawug.org/>
[un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless

Reply via email to