93 mbps direct to modem, 30 mbps wireless N?

[MBrown]

I am a noob when it comes to networking, but I just hooked one of my computers directly up to my modem wired and got 93 mbps download, but when I connect through connect through wireless N I get 30 mbps max. Wireless N spec is 300 mbps so why doesn't it hit the same speeds as the wired connection? Yes I am running 5Ghz.

 

[PliotronX]

Might be a dumb question but are you running Windows? What data rate does Windows report? A result of 30Mbps is low when it reports 150Mbps and high for 75Mbps. I've never gotten 300Mbps to work reliably with a multitude of devices because if there is more than one other AP in the area or interference source, the channel bonding does not work. Not nearly as well as the Atheros "turbo" 108Mbps back in the 11g days. I think 30-40Mbps is about the max an N can muster in the real-world.

Also JFYI 5GHz is not 802.11n, it is 802.11a. The "750Mbps" routers use both 2.4GHz and 5GHz to boost performance but IME 5GHz is hit or miss. It's less susceptible to interference from common sources but has a hard time penetrating walls worth a damn.

 

[smitbret]

Might be a dumb question but are you running Windows? What data rate does Windows report? A result of 30Mbps is low when it reports 150Mbps and high for 75Mbps. I've never gotten 300Mbps to work reliably with a multitude of devices because if there is more than one other AP in the area or interference source, the channel bonding does not work. Not nearly as well as the Atheros "turbo" 108Mbps back in the 11g days. I think 30-40Mbps is about the max an N can muster in the real-world.

Also JFYI 5GHz is not 802.11n, it is 802.11a. The "750Mbps" routers use both 2.4GHz and 5GHz to boost performance but IME 5GHz is hit or miss. It's less susceptible to interference from common sources but has a hard time penetrating walls worth a damn.

What?

I am thoroughly confused by your response to OP's question. A couple of items just seem off the wall and/or incorrect but there's enough wiggle room that I entertain the idea that I may just me misinterpreting.

5GHz is most certainly an n standard. I agree wiht you that it's penetration limitations mean that I really don't use it for anything that roams, just use it between bridges, etc.

750Mbps routers do not use both bands to boost performance. They are dual band routers that are literally running 2 separate networks, a 2.4GHz @ 300Mbps and a 5GHz @ 450MBps. You will be limited to those maximum connection speeds, respectively depending on which network you connect.

I do agre with your assessments though that 30-40Mbps is probably a good real world expectation for a mobile device.

 

[imagoon]

Wireless is a shared lossy medium. It is hard to determine how well it will work without knowing the ambient conditions.

Also "300mbps" (and the above mentioned 750mbps) in wireless is just symbol / signalling rate. Wireless N is normally only going to get you about 60% of the symboling rate in perfect conditions. So 160mbps is a realistic perfect world rate (one way AP => device.) assuming ideal conditions. Interference, multiple direction protocols, quality of the radio will all play a part in what the real world connection is.

 

[PliotronX]

What?

I am thoroughly confused by your response to OP's question. A couple of items just seem off the wall and/or incorrect but there's enough wiggle room that I entertain the idea that I may just me misinterpreting.

5GHz is most certainly an n standard. I agree wiht you that it's penetration limitations mean that I really don't use it for anything that roams, just use it between bridges, etc.

750Mbps routers do not use both bands to boost performance. They are dual band routers that are literally running 2 separate networks, a 2.4GHz @ 300Mbps and a 5GHz @ 450MBps. You will be limited to those maximum connection speeds, respectively depending on which network you connect.

I do agre with your assessments though that 30-40Mbps is probably a good real world expectation for a mobile device.

Damn, well I had mistaken it because 5GHz is so underutilized and I just see horrible range with it that it's almost not in the zeitgeist.

So how can router manufacturers get away with advertising 750Mbps? Isn't someone going to get them to stop advertising insane speeds like 1Gig with 11ac? Have you seen any manufacturer brave enough to forgo the "up to" speeds and report what users would actually see? I've gotten bothered about wireless speeds just about as much as how to do stuff in Windows 8, I'm pretty tired of it.

 

[azazel1024]

Because it kind of sort of is.

A 750Mbps router is likely 300Mbps 2.4GHz and 450Mbps on 5GHz and is concurent dual band. IE both 5GHz and 2.4GHz radios can send/receive at the same time.

A handful of routers/bridges/access points actually allow you to bridge a connection using both bands at the same time...which theoretically could give you a real 750Mbps bridge.

Then you have to knock off 802.11 overhead. In the case of 11n it is something in the ballpark of 40% when you account for error correction data, beacon interval, preamble, CTS/RTS, etc. 11g it was around 60%. 11ac its around 40% like 11n...but in part because it is 5GHz you are a bit more likely to run in to lower signal issues.

In reality, close to an access point with little noise, you could get around 750Mbps true throughput to the router if you have a 2.4GHz and a 5GHz client. Of course only about 60% of that is useful for package data (IE what YOU, the user cares about).

Problem here is, few if any clients are concurrent dual band.

The only time I think wireless manufacturers are truely deciteful is when they are selling a product like a 600Mbps 2.4/5Ghz concurrent dual band router...that only has 10/100 ports on it, as it can't possibly exceed 100Mbps, if that.

 

[smitbret]

Damn, well I had mistaken it because 5GHz is so underutilized and I just see horrible range with it that it's almost not in the zeitgeist.

So how can router manufacturers get away with advertising 750Mbps? Isn't someone going to get them to stop advertising insane speeds like 1Gig with 11ac? Have you seen any manufacturer brave enough to forgo the "up to" speeds and report what users would actually see? I've gotten bothered about wireless speeds just about as much as how to do stuff in Windows 8, I'm pretty tired of it.

They get away with it because the router can be providing that much RATED througput at a time if you have multiple devices. If you have a 450Mbps client on the 5GHz band and a second client at 300Mbps on the 2.4GHz, then the router will have 750Mbps worth of connections running simultaneously.

That's their loophole.

What's silly is that if they did the same thing with the wired connections, they could make some REALLY insane high numbers, but they don't. If they just applied the same formula to a 4 port gigabit router then they could advertise:

8Gbps!!!!!!!

I don't know how that keeps slipping past marketing

 

[JackMDS]

Because it kind of sort of is.

A 750Mbps router is likely 300Mbps 2.4GHz and 450Mbps on 5GHz and is concurent dual band. IE both 5GHz and 2.4GHz radios can send/receive at the same time.

A handful of routers/bridges/access points actually allow you to bridge a connection using both bands at the same time...which theoretically could give you a real 750Mbps bridge.

Then you have to knock off 802.11 overhead. In the case of 11n it is something in the ballpark of 40% when you account for error correction data, beacon interval, preamble, CTS/RTS, etc. 11g it was around 60%. 11ac its around 40% like 11n...but in part because it is 5GHz you are a bit more likely to run in to lower signal issues.

In reality, close to an access point with little noise, you could get around 750Mbps true throughput to the router if you have a 2.4GHz and a 5GHz client. Of course only about 60% of that is useful for package data (IE what YOU, the user cares about).

Problem here is, few if any clients are concurrent dual band.

The only time I think wireless manufacturers are truely deciteful is when they are selling a product like a 600Mbps 2.4/5Ghz concurrent dual band router...that only has 10/100 ports on it, as it can't possibly exceed 100Mbps, if that.

+1 :thumbsup:
==============

Main reason for the confusion is Consumers ignorance/wishful thinking.

The specs of Wireless are combo of technical ratings Not functional rating.

Like putting a Chipset that is rated 300Mb/sec. but in the totality of the card design, heat dissipation, and Antenna, ends up functioning much slower regardless of what the OS reads in the drivers table. The specs. also assume that the source and the client are in the same small room in an electronically isolated environment (no interference what so ever) and transferring short burst of simple light packets.

In functional reality No Body has such an environment.

 

[imagoon]

They get away with it because the router can be providing that much RATED througput at a time if you have multiple devices. If you have a 450Mbps client on the 5GHz band and a second client at 300Mbps on the 2.4GHz, then the router will have 750Mbps worth of connections running simultaneously.

That's their loophole.

What's silly is that if they did the same thing with the wired connections, they could make some REALLY insane high numbers, but they don't. If they just applied the same formula to a 4 port gigabit router then they could advertise:

8Gbps!!!!!!!

I don't know how that keeps slipping past marketing

Not quite. They list the symbol rates. Symbol rates are only partly related to the real datarates. For example: G is 6 bits of every 8 bit symbol that is sent. Add in error correction (this is above symbols in the wireless stack), protocol overhead etc and you arrive at the "40%" rule of thumb. It is also fairly odd way to do it since the symbol data rate is rarely used. They normally call the symbol rate (Sym/sec.)

You can do the same thing on Gigabit but it doesn't sound "as cool." 100mbit and 1000mbit have the same symbol rate (per transmitter). Gigabit only has 4 of them + 8 symbol bits to 16 bits of data. Which is "cooler (marketing)" 1000mbits or 500msymbols or "500 megabits?"