[174971] in North American Network Operators' Group
Re: Marriott wifi blocking
daemon@ATHENA.MIT.EDU (Owen DeLong)
Tue Oct 7 08:06:41 2014
X-Original-To: nanog@nanog.org
From: Owen DeLong <owen@delong.com>
In-Reply-To: <07A070D3-C36A-48F2-A40D-7F57E28341BA@bloomcounty.org>
Date: Tue, 7 Oct 2014 05:05:48 -0700
To: Clay Fiske <clay@bloomcounty.org>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
On Oct 6, 2014, at 11:53 AM, Clay Fiske <clay@bloomcounty.org> wrote:
>=20
> On Oct 6, 2014, at 8:41 AM, Owen DeLong <owen@delong.com> wrote:
>=20
>>=20
>> Actually, in multiple situations, the FCC has stated that you are =
responsible
>> when deploying a new unlicensed transmitter to insure that it is =
deployed in
>> such a way that it will not cause harmful interference to existing =
operations.
>=20
> I recognize that you were making this statement in the context of =
colliding SSIDs, but to me this could be an interesting point in another =
way.
>=20
> Suppose from Marriott=92s perspective that your personal wifi network =
is interfering with the throughput of their existing network. After all, =
if you fire up your personal AP, with a non-colliding SSID, and start =
downloading multi-GB files, that=92s bound to impact[1] anything else =
using that channel. While there are at least a few non-overlapping =
channels on most wifi networks, if Marriott(=92s third party network =
operators) had any sense they likely would have situated their APs and =
channels to provide the most range with the least amount of frequency =
overlap. Now here your personal AP on one of those channels consuming =
enough of its bandwidth to significantly degrade performance for anyone =
else, and they may not have access to (or usable signal strength or =
bandwidth on) another channel from their hotel room.
The FCC has specifically stated that sharing of the spectrum bandwidth =
in this manner is not considered =93harmful interference=94 in at least =
a few rulings. This is the =93normal and expected result of deployment =
of multiple networks onto limited spectrum=94.
> During a big convention for example, the hotel network is probably at =
its busiest while the number of guests using personal APs is likely also =
at its peak. This may be a stickier case, as no one user is causing the =
issue but one could make the case that, in aggregate, they are very much =
interfering with existing operations.
Yes, but not in a manner the FCC fits into the definition of =93harmful =
interference=94 under 15.3 and/or 15.5.
> There are probably a couple of different angles to consider, but I=92m =
thinking in terms of the =93first come, first served=94 concept. At what =
point is the extra bandwidth consumed by your personal wifi network =
considered to be harmfully interfering with an existing network?
It isn=92t (unless you run afoul of 47CFR333 and are consuming bandwidth =
for the sole purpose of denying it to others).
> FWIW I am not defending Marriott=92s actions, nor even positing that =
this was the reason for them. I just want to gain understanding.
Yep. Understood. Hope the above helps.
> -c
>=20
> [1] This is of course assuming you=92re getting decent throughput from =
your 3G/4G provider=92s network. But even though it=92s almost certainly =
slower than wifi it=92s probably generating enough packets in a =
collision-based medium to impact other flows.
Actually, I usually get better 4G service on my LTE devices than I get =
from most hotel WiFi networks. It=92s one of the reasons I wish Apple =
would let me choose the interface preference order rather than locking =
me to =93if Wifi is on and can find an AP, then I won=92t use LTE=94.
Owen
