How I doubled my Internet speed with OpenWRT
msol.io
msol.io
I did tried dd-wrt for this same purpose, the interface is much better, it is very flexible as openwrt, but I couldn't put my wifi on client and AP mode at the same time and I needed to connect to one of my WAN using wifi because that modem is not close enough yet.
if you have older x86 hardware, it's hard to top pfsense.
I bought it for AUD$50. The closest router in terms of price + power was a TP-Link for around AUD$70 and it still was behind on features.
because you know, capitalism and stuff.
[1]: http://www.tp-link.com/resources/images/products/large/TL-WD...
BUT OP will only double his bandwidth if he is purchasing a rate less than or equal to the actual realized throughput of his WiFi connection, somewhere from 50-90Mbps.
Additionally, all Cable modems have some kind of throttling because the speeds offered by ISPs (30x5, 100x8, etc.) are not evenly divisible by the per-channel rate of DOCSIS 2.0/3.0 (38Mbps Down, 27Mbps up).
And then there's this: http://www.dslreports.com/forum/r29743167-Signals-Report-16-...
A piece of coax has a finite amount of bandwidth it can carry. As modulation schemes get better (QAM256) you can do more with less.
To say DOCCIS 3 is limitless is not taking into consideration the hardware and frequency constraints. The more frequency you allocate to Internet the less is available for other programming (TV).
If you have a Motorola cable modem go to: 192.168.100.1 to get to the web UI of your Surfboard. Click on the "Signal" tab at the top. My guess is the majority of people will see 4 x bonded down. Probably 2-3 bonded up using a lesser QAM as well.
Not sure why all the hate - I'm a network engineer who's dealt with these networks and a lot of what's been posted here is incorrect.
I'm curious how much faith to put in the speedtest numbers. Does Comcast detect speedtest, or is that what I'd really get elsewhere as well?
Here's a list: http://www.glasvezel.nu/speedtest-servers-100mb-1000mb-bin
From my experience with low throughput 2-4mbps lines I'm deeply sceptical about claims like that. Throughput will generally be OK...but if you want stable latency...good luck. The second someone else is on the same line goes flakely unless its a high powered line (fibre etc)
I blogged about my setup at http://www.sajalkayan.com/post/fun-with-mptcp.html
I actually use this method to use school WiFi anonymously (or rather, as someone else).
It's how eduroam works, and that works fairly flawlessly (provided the routers have enough bandwidth).
"Passpoint automates that entire process, enabling a seamless connection between hotspot networks and mobile devices, all while delivering the highest WPA2™ security. Passpoint is enabling a more cellular-like experience when connecting to Wi-Fi networks."
If you want the fastest speeds on Comcast, pick up a DOCSIS 3 Motorola Surfboard modem. I'm paying for 100/20. This is what I get with my surfboard http://www.speedtest.net/my-result/4251868226
No.
I feel like this post should be longer, but I can't think else to say. The answer to your question is: No.
The only way it would increase bandwidth is if it evades artificial throttles. Having two connections to the same (neighborhood) pipe seems useless.
Possibility two: placebo effect.
UPC only makes a small amount of bandwidth available on the 'Wi Free' (UPC's version of Xfinity) network, seems to be about 2.5mbps down and 0.5mbps up. In my country (Ireland), the minimum package you need to get 'Wi-Free' is 120mbps down, so It's not really worth the effort for an extra 1% bandwidth.
Well obviously. They give you a specific bandwidth for your house, you can buy more or less bandwidth if you want.
> Having two connections to the same (neighborhood) pipe seems useless.
The neighborhood pipe can probably do 100 to 1000 times the bandwidth to each individual house. So how is that useless?
Plus with xfinitywifi they allocate double bandwidth to the house, so the wifi bandwidth doesn't slow down the purchased bandwidth. (The cable modem can easily handle 10 times the bandwidth typically allocated to it.)
Apart from that, my understanding is that Comcast routers broadcast your network and also broadcast this comcast hotspot network, and they are metered (and presumably capped) independently. So with this hack you would use only your equipment, you're just taking advantage of both networks that are being broadcast by your router.
Check the article again. The author is connecting his router to his neighbor's hotspot.
I don't have a lot of confidence in Comcast to have set it up so that connections to the xfinitywifi network do not impact the regular, private usage. It's just not in that company's style to think proactively and conscientiously about its customers.
I don't think there is such a thing as a comcast account without a data cap.
Note: Enforcement of the 250GB data consumption threshold is currently suspended.
(427GB so far this month...)
That's besides the point that monthly caps is a completely nasty rip off. Luckily I never had one. Are they common in Comcast network?
I think what he's trying to say is that as an individual using the xfinitywifi network the traffic counts towards your data cap. As the person hosting the access point, other people using it does not count towards your cap.
Ah, that makes more sense.
Does anyone know if this OpenWRT works? I would love to push my speed to its limits. Comcast pretty much has a monopoly in my city and they are not very helpful.
Wifi is extremely finicky, due to interference and the FCC's limits on legal transmit power levels. But you should be able to get same-room speeds of at least 180 megabits if you have a decent router and a client that supports it. The key things are you need to use a 5 GHz radio and (for distance) channel numbers >100. Many ISPs still give out 2.4 GHz-only radios, so performance will suck.
If you want to get a better Wifi connection, get a good router and plug it in to the one installed by Comcast.
If the grocery store makes a big deal out of having staff to carry groceries to your car, and how they sell 120 groceries, they had better be able to carry all 120 to cars that support it.
Either buy a Cat5e/Cat6 cable or change your wifi access point to use 802.11n. I am connected to my 2007 Airport Extreme at 300mbps right now using 802.11n.
This has more information on cat5 vs cat5e vs cat6 and their compatibility with different speeds / modes: http://serverfault.com/questions/107172/what-is-the-actual-d...
TL;DR almost any Ethernet cable you buy today will be cat5e or better, which will work just fine. Really old cat5 is technically compatible and should work too, but due to it having slacker tolerances for e.g. crosstalk, it might not work over e.g. longer runs or difficult situations (cable ran next to fluorescent lights, etc).
EDIT: If you want to know what speed your link-level has established, your OS should provide this info. netstat -e on windows or netstat -i on OSX / Linux should show the connection speed. In windows, you could also do WinKey+ r -> ncpa.cpl -> right-click active internet connection -> connection details IIRC.
You could of course make the argument that you bought some combined cable modem/router/wifi AP box from them and it should "just work" - but you're just going to talk until you're blue in the face. Simply think of them as a pipe provider (not your IT support) and you'll be a lot happier.
What they provide is in general crap, and costing you money you can save off the monthly rental fee by providing your own equipment. This will also generally result in better performance.
I suggest buying a Surfboard 6183 (or lesser, depending on your area), and the wireless router of your choice. I recommend for simplicity sake the Apple Airport extreme or any quality 802.11n (or 802.11ac if your laptop supports it) router. Your network should look something like Cable from wall -> DOCSIS Modem -> Your wifi router. The only time you should be calling comcast is when the actual cable modem cannot obtain a lease and/or internet is in general failing.
Relying on Comcast to manage your CPE is just a plan for disaster. If your wired tests are getting 100Mbps, then the problem is on your end.
I imagine the less technically-inclined would have given up a lot earlier in the process.
In fact I've noticed conveniently broken web pages from multiple large corporations. I've ran into this issue when attempting to access privacy policies and other disclosures.
There certainly may be something nefarious about this. Comcast could certainly get away with it.
I set up service with my own equipment from the start, and after all the account setup I received a dozen notices that all these wifi points were available. It was never specified as a "must have a Comcast router" condition.
http://customer.comcast.com/help-and-support/internet/disabl...
Also, because your router is a routed client of the xfinity wifi network, I'd imagine there'd be a big increase in latency. It'd be interesting to see the before/after speed test results.
My service is 3Mbps. My modem & my hardline are both capable of much more, but I only pay for 3Mbps so I only get 3Mbps. I think this trick is basically allowing the author to tap into the unallocated extra modem & line capacity that is currently used to feed xfinitywifi.
The best way to improve consumer Internet connection is to get a fast router that can route fast in hardware. I'm always amazed people think a SOHO device doing WiFi, NAT, DHCP, DNS, etc. on gimped hardware is "fast". The majority of time it's not and real improvements can be realized with dedicates hardware. Meaning that until you split service off from routing using cheap, consumer SOHO gear, will most always be the bottleneck.
Again, he's not doubling his bandwidth based on the above and a whole host of other reasons beyond these.
For those who try I'd say post true bandwitdh and latency test comparisons using multiple sessions that are shown using both links. Its actually pretty funny to me, having been in network engineering for well over a decade, that people are so passionate that they're sticking it to Comcast and "doubling" bandwidth with nothing to back it up. But, whatever floats your boat.
> Does the new Home Hotspot impact my Internet speeds or data usage?
> The broadband connection to your home will be unaffected by the new feature.
Doesn't sound consistent with the prioritization claim since prioritization still implies the speed may be affected by the guest network. If you're right, then that's kind of a bold (misleading) thing for Comcast to say in their FAQ.
If the author is doing this by using the neighbor's WiFi (as mentioned in some other comments), then I don't see how this has anything to do with XFINITY WiFi specifically, or any loopholes - stealing bandwidth from a neighbor could be done (in theory) with any open/accessible WiFi connection of a neighbor.
DOCSIS (not DOCCIS) can do a Gb/s, and that's just to one modem, not the neighborhood as a whole.
I just bought a TP-Link TL-WR841N and it has no problem maxing out my 30mbps connection, even with some extra LAN traffic.
He isn't connecting to the same modem twice.
You seem to be dripping misplaced condescension all over the floor - would you like me to fetch a mop?
When you rent a cable modem/router combo from Comcast (as one of my nearby neighbors apparently does)
The article definitely implies that he's connecting to his neighbour's Comcast router.
Think of it this way. When you stream a movie from Netflix it is one TCP session. Not 5, 10, etc. So if you load balance to your neighbor your stream will be worse by default. This is not equal cost load balancing, this is a hack that isn't what it seems and there are a lot of people commenting that have, apparently, little to no knowledge about basic network fundamentals.
Try it and prove you've doubled your bandwidth. There are tens of logical reasons why this doesn't work.
Like others have pointed out, Linux doesn't load balance a single connection over multiple WANs, so as long as the Netflix stream gets on the best link, it'll be better since it'll have to compete with fewer connections (since some will be routed over the other link).
4 lane highway vs 1 lane dirt road. While combining the two makes the number of paths greater the paths themselves are not equal (my equal cost load balancing remark). So, if the router is not taking link cost into account, which it is not, you'll actually have worse performance over all of your connections because the router will try to balance them equally on session start. This means that 50% of the time you start out going to your neighbors connection and fail or go slow. It's actually more advantageous to wait for a spot on the wide, fast link than it would be to take the alternate path.
If the router was taking into account link reliability and speed (overall cost) it would only give a small percentage of connections to the worse link and only if the main link was saturated from a bandwidth perspective. None of this is happening by round robin load balancing which means more of your connections are worse. This is why routing protocols that have these features exist.
Your comments are clearly not kind or necessary, and not all of them are even true. You might wish to rethink your commenting strategy, because you're clearly smart and I would rather that your opinion not be lost entirely.
> When you rent a cable modem/router combo from Comcast (as one of my nearby neighbors apparently does)
+----------------------+
| |
+-----------> | Neighbor's router |
| | |
+-----------+ +------------------+ | +----------------------+
| | | | +
| OP's PC +----> | Buffalo Router | +----------------------+
| | | | + | |
+-----------+ +------------------+ | | OP Comcast connection|
+-----------> | |
+----------------------+
I'm a big fan of my TP-Link WDR4300. I can easily max out my 250/60Mbps uplink while it's also doing NAT and some light firewalling, in software. It also runs all of my IPAM (DHCP/DNS). Additionally, it also runs OpenVPN (and can do ~20Mbps of encrypted bandwidth) and a BGP session (using Quagga) over that VPN to my local hackerspace. All in a single OpenWRT device that's sub $70.
Is it equivalent in performance to a hardware router? Of course not. But these start at a few thousand dollars (even Cisco ASA and Juniper SRX class hardware does its routing in software...).
I like the idea but I wonder how it performs on many use cases (like Skype or online gaming).
from http://wiki.openwrt.org/doc/howto/mwan3
Linux outgoing network traffic load-balancing is performed on a per-IP connection basis – it is not channel-bonding, where a single connection (e.g. a single download) will use multiple WAN connections simultaneously
I can still easily imagine a conflict where your signaling IP and your media IP are different in VoIP, but binding connections to IP addresses is pretty reasonable.
I don't see anything particularly noteworthy here. He just happens to have a very specific setup and is leveraging it.
I live in a typical neighborhood in Chicago; houses are spaced apart enough to where you don't get a ton of wifi overlap, but enough to see a handful of your neighbors networks. I noticed the first "XfinityWifi" network about mid/late last year and now will see between 2 - 3 from my house. I found this post interesting because of how he is leveraging these, relatively, open connections.
"When you rent a cable modem/router combo from Comcast (as one of my nearby neighbors apparently does)"
It may be illegal:
http://en.wikipedia.org/wiki/Legality_of_piggybacking
Comcast is not on the EFF's list of Wireless Friendly ISP's (referenced in the Wiki article):
https://www.eff.org/pages/wireless-friendly-isps
Even if it were: http://compnetworking.about.com/od/wirelessfaqs/f/legal_free...
"Using a neighbor's wireless access point may not be legal even with their permission. "
https://www.techdirt.com/articles/20141208/13222529362/comca...
It has been proven that in many places at least in the US, it is illegal and in some cases is a felony.
Edit: Further, it is not a good idea to connect to any known SSID unless you know for a fact that it can be trusted:
http://arstechnica.com/security/2014/06/free-wi-fi-from-xfin...
Your neighbor might be hacking you while you think you are legally using their hardware.
Anything else including what the author describes, using his neighbor's equipment without their permission even if he thinks Comcast has the legal authority to authorize it (and documentation in the links I provided challenges that notion) then it's not for me to say; I'm not a lawyer.