wireless faster than wired connection??

"VanguardLH" schreef in bericht
...
Linea Recta wrote:

I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the notebook
by
WiFi scores faster than the PC.
Something wrong with the PC configuration?

How much faster? A few Mbps is within the deviation of testing. You
need to discard the first test, run the test 5 (10 is better) more
times, and then average those last 5 (or 10) tests to get a value to
compare against the other computer. What time of day did you use the
speed test site? Hit them when less busy, like when it's mid-morning
for them.


Sorry for the delay, afraid I have to do things simultaneously here.
Time of the day, good point, I haven't thought about that. I suppose that
explains the difference mainly. I use the notebook at dayv time mainly,
while using the PC usually at night.



Did the speedtest.net test (or whatever site you used that you did not
identify) show speeds above 10 Mbps? If so, is the NIC (card or
onboard) in the PC capable of speeds higher than 10 Mbps? That you have
Windows 7 installed on a computer doesn't necessitate that it is a
relatively new computer. It could be an old computer whose NIC can only
get up to 10 Mbps (a max that is not achieved for sustained transfers).


Both computers are relatively old.



Is the router configured with QoS (Quality of Service) to give priority
to wifi traffic over cabled Ethernet traffic?


I did not (yet) have time to explore all router setings extensively, apart
from channel setting (auto).




How long is the CAT-5 cable from PC to the router? Did you try a
different (and shorter) cable? Did you try plugging the wired computer
into a different RJ-45 LAN port on the router?


The cable is about 1 meter long. I can't imagine this is too long.
The modem/router is a FritzBox 7360.



Did you use a speed test page provided by your ISP (so you remain
in-network with your ISP for speed testing) or did you bounce out and
back in by using an external speed test site?

External, see data below.



When you ran the speed test, did the web site connect to the SAME test
center for both tests?


Obviously yes.


If 2 different sites were used then there are
different nodes in each hop between you and that site so latency and
responsiveness will vary. Did you try testing to a different site than
the default one they pre-selected for you? The speed test site rarely
has control over what hosts are in the hops between it and the speed
test site (the speed site doesn't operate its own data centers so
borrows hosts from elsewhere to do the tests). Physically close is not
always a good choice. A host that is next door might require you to go
over your ISP's network to a backbone provider off some hundreds of
miles away to reach another city to bounce back to that next-door host
that is using a different ISP. Speedtest.net likes to pick a test site
in a western suburb from my city along as the default. I can also pick
the state university in the same city, a city over a hundred miles in
the next state, or a test site in Chicago which is the same physical
location as my ISP's backbone provider. The university and Chicago give
the most consistent timings. The next-state test host has a bit more
variation. The suburb site varies a lot. Since I cannot perform a
traceroute from the speed test site to the actual speed test hosts, I
can't tell if the suburb site has more hops and what they are compared
to the other test sites.

What security software might you have installed on the PC that is not
installed on the wifi notebook?


Security software the same on both computers.


All security interrogation incurs time
to inspect hence adds delay. Did you test your PC by rebooting it into
Windows' safe mode with networking to ensure that nothing you load on
the PC is causing lag with network traffic?

No.

Below some measurements I made yesterday.



speedtest site:
http://www.speedtest.nl/

vrijdag 14 november 2014
local time 22:15

PC measurements
--------------------------------

Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 715.1 KByte/sec
(=5721 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 164.5 KByte/sec
(=1316 kbps)

------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 747.7 KByte/sec
(=5982 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 172.5 KByte/sec
(=1380 kbps)

-------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 1100.6 KByte/sec
(=8805 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 167.5 KByte/sec
(=1340 kbps)

-------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 680.7 KByte/sec
(=5446 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 170.7 KByte/sec
(=1366 kbps)

-------------------------------------------------------------------------------------

Notebook measurements:


Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 809.7 KByte/sec
(=6478 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 137.3 KByte/sec
(=1098 kbps)

---------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 1059.3 KByte/sec
(=8474 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 151.3 KByte/sec
(=1210 kbps)

----------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 1168.9 KByte/sec
(=9351 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 152.7 KByte/sec
(=1222 kbps)

------------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 731.6 KByte/sec
(=5853 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 231.9 KByte/sec
(=1855 kbps)

----------------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 1029.3 KByte/sec
(=8234 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 228.6 KByte/sec
(=1829 kbps)

---------------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 1199.2 KByte/sec
(=9594 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 151.7 KByte/sec
(=1214 kbps)

--------------------------------------


NB today:
(all notebook measurements)


zaterdag 15 november 2014
12:05 local


Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 3738.7 KByte/sec
(=29910 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 376.5 KByte/sec
(=3012 kbps)

-------------------------------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 3794.5 KByte/sec
(=30356 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 335.1 KByte/sec
(=2681 kbps)

-----------------------------------
Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding) vanaf
een server in Nederland Klaar Gegevens opgeslagen ? 3259.4 KByte/sec
(=26075 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de upload
snelheid te bepalen Klaar Gegevens opgeslagen ? 323.2 KByte/sec
(=2586 kbps)

--------------------------------------




--


|\ /|
| \/ |@rk
\../
\/os

Linea Recta wrote:

FritzBox 7360

Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 3794.5 KByte/sec
(=30356 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 335.1 KByte/sec
(=2681 kbps)


You get some nice speeds there. That's high enough for it to
not be ADSL2+, but be VDSL.

I would keep the interconnect and equipment end constant,
and check for time-of-day variation. Maybe there are
simply too many high-bandwidth VDSL users on your DSLAM,
all downloading (or watching IPTV) at night.

Paul

On 11/14/2014 6:32 PM, Leala wrote:
On 14-Nov-2014 18:06, Roger Mills wrote:
On 14/11/2014 18:41, Good Guy wrote:
On 14/11/2014 18:26, Linea Recta wrote:
I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?




Speed of light is still the fastest and so wired connection cannot
compete with Wi-fi in terms of speed. However, wired connection is much
more robust than the wi-fi and so if you want to avoid any corruption
when downloading large files then always use a wired connection.


I can't work whether you're joking or just stupid!

I hope he's joking otherwise he doesn't know what he's talking about.


My observation is that it is the later.

"Paul" schreef in bericht
...
Linea Recta wrote:

FritzBox 7360

Download
Download van een bestand (128KB ~ 1MB afhankelijk van uw verbinding)
vanaf een server in Nederland Klaar Gegevens opgeslagen ? 3794.5
KByte/sec
(=30356 kbps)
Upload
Upload van gegenereerde gegevens naar een server in Nederland om de
upload snelheid te bepalen Klaar Gegevens opgeslagen ? 335.1 KByte/sec
(=2681 kbps)


You get some nice speeds there. That's high enough for it to
not be ADSL2+, but be VDSL.

I would keep the interconnect and equipment end constant,
and check for time-of-day variation. Maybe there are
simply too many high-bandwidth VDSL users on your DSLAM,
all downloading (or watching IPTV) at night.



OK thanks Paul. I suppose the 'time of the day' answers the question. My
fault this didn't occur to me in the first place.



--


|\ /|
| \/ |@rk
\../
\/os

Linea Recta wrote:

"VanguardLH"

is the NIC (card or onboard) in the PC capable of speeds higher than
10 Mbps? It could be an old computer whose NIC can only get up to
10 Mbps (a max that is not achieved for sustained transfers).

Both computers are relatively old.

Okay, but what are the max speeds for the NIC in each computer?

What speeds are being reported by the unidentified speed test site?

Did you use a speed test page provided by your ISP (so you remain
in-network with your ISP for speed testing) or did you bounce out and
back in by using an external speed test site?

External, see data below.

Check if your ISP provides their own speed test page. If you cannot
find it wandering around their site, use their tech chat to ask about
speed testing but staying within their network.

When you ran the speed test, did the web site connect to the SAME test
center for both tests?

Obviously yes.

Not obvious. When I revisit speedtest.net, they often pick a different
test host. There are a couple dozen in my area and which ones they show
and which one they pick by default might have to do with some algorithm
to rotate through those that are up (since those that are down are not
usable) along with how busy they are. They don't operate their own test
hosts. They get them from other sites willing to do the testing.

Did you test your PC by rebooting it into Windows' safe mode with
networking to ensure that nothing you load on the PC is causing lag
with network traffic?

No.

Same software can have different configurations on different hosts.
Different hardware on the hosts can alter the performance of the same
software.

Below some measurements I made yesterday.

http://www.speedtest.nl/

vrijdag 14 november 2014
local time 22:15

PC measurements

Download = 5721 kbps
Upload = 1316 kbps

Download = 5982 kbps
Upload = 1380 kbps

Download = 8805 kbps
Upload = 1340 kbps

Download = 5446 kbps
Upload = 1366 kbps

Notebook measurements:

Download = 6478 kbps
Upload = 1098 kbps

Download = 8474 kbps
Upload = 1210 kbps

Download = 9351 kbps
Upload = 1222 kbps

Download = 5853 kbps
Upload = 1855 kbps

Download = 8234 kbps
Upload = 1829 kbps

Download = 9594 kbps
Upload = 1214 kbps

There is a LOT of variance in the downstream measurement from the test
host they or you selected (Nederland Klaar Gegevens opgeslage). For the
PC measurements, the average is 6.5 Mbps but the variance is huge at 1.4
Mbps (but the data count is a mere 4). For the notebook measurements,
the average is 8 Mbps but the variance is also huge at 1.4 Mbps (with
only 6 data points). Values are too erratic to make a decision as to
which computer is faster; however, they roam in the same ballpark so I
can't, with the data at hand, say that one is faster than the other.

That is NOT a good host (or the route to it) for speed testing. Values
are too far ranged. That host is worthless for knowing what speed you
have (with your ISP since anything outside them is not controlled by
them). You need to use something that produces consistent measurements.

Try a different test host than what they pick for you. If you get
similarly far flung measurements, try a different host. As I mentioned,
physically close is not always a good choice to speed measurement. See
if your own ISP has their own speed test page.

For comparison, I used my ISP's speed test page and got the following
values (in Mbps) in 5 tests:

up = 18.76 down = 12.26
up = 18.77 down = 12.15
up = 18.75 down = 12.07
up = 18.73 down = 12.16
up = 18.75 down = 12.19

Notice how much tighter are those measurements. The average is 18.8
with a meager 0.013 variance. I then did 5 tests at speedtest.net.
Their measurement is way off (typical) because I do *not* have the 50
Mbps service tier from my ISP. I have the base plan of 25 Mbps (I got
lower measurements probably because I measured at 10 AM local which
might be busy with typical users awake during the day). However, notice
the consistency of the Speedtest measurements for the test host that I
selected (which was also their default site for today in my area).

up = 59.31 down = 12.15
up = 59.30 down = 12.12
up = 59.36 down = 12.15
up = 59.30 down = 12.13
up = 59.33 down = 12.20

The average is 59.3 with a variance of only 0.023. I suspect SpeedTest
rates much higher due to the PowerBoost feature of my ISP versus my
ISP's own test page nullifies that feature. That is, if SpeedTest's
files are small (under 20 MB) then my ISP will increase bandwidth during
that transfer (or the first 20 MB if the file is larger). All I could
find at the SpeedTest site was they download "small binary files".
"Small" is not defined.

You need to use a test host (that does the actual benchmarking) that
provides far more consistent measurements than the one you picked as
shown in your reply. Your ISP's own speed test page would be a likely
choice; else, you have to hunt around evaluating the hosts that
SpeedTest lists to see which one is good enough for reliable benchmarks.

Chuck Anderson wrote:

Linea Recta wrote:
I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?

Could be. My wired Ethernet NIC card limits me to a 100MB bandwidth (not
Gigabit). My wireless card, however, will do up to 300MB.

My Wireless be faster (vroooom) than wired. ô¿Ô¬

But what bandwidth service tier do you pay from your ISP? Having a
bigger pipe on your end doesn't force their pipe to get bigger. If they
provide only 25 Mbps for what you pay them, you having a 100 Mbps or 1
Gbps pipe won't magically force them to give you any more than the 25
Mbps that you pay for.

Chuck Anderson wrote:
Linea Recta wrote:
I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?

Could be. My wired Ethernet NIC card limits me to a 100MB bandwidth (not
Gigabit). My wireless card, however, will do up to 300MB.

My Wireless be faster (vroooom) than wired. ô¿Ô¬


The difference though, is that wired cards tend to work at "wire speed".
If the NIC claims to be 100Mbit/sec (100BT), then it actually runs the
wire at that speed, with tiny gaps for inter-packet spacing.

The 802.11n theoretical 300Mbit/sec relies on a number of conditions
to be true. And on typical equipment and sites, they're not true. To
start with, no Wifi negotiates to the theoretical value, it always
seems to be one step lower (so like half maybe, best practical case).

That's why the FritzBox 7360 VOIP router here, benches at

http://www.zdnet.com/avm-fritzbox-73...rs-4010026506/

"they delivered fairly similar throughput of around 30-35Mbps at 1meter"

which is a far cry from the 300Mbit/sec theoretical best case. You would
really have to look at a large number of their reviews, to get a sense
of whether they're screwing up somewhere. For example, one lab that
was careful in their testing, used an RF absorptive mat near the test
area, and got an increase in performance. Their 802.11n has probably
reverted to 20MHz channel spacing, because it "sniffed" legacy Wifi
in the vicinity. The number still seems to be on the low side though.

The FritzBox uses "crazy looking" antennas, just some soldered inverted
U-links. I don't remember seeing any other antenna implementations
that look like that :-)

*******

If you want to benchmark LAN performance, to compare the
wired NIC path to the Wifi, you could use PCATTCP. One machine
on your LAN functions as the transmitter, the other machine as
the receiver, and it does a short test. That's slightly better
than setting up RAMDisks and using Windows File Sharing for testing :-)

http://www.pcausa.com/Utilities/ttcpdown1.htm

*******

In terms of winning bragging contests, 802.11ad will do that for you.

http://www.extremetech.com/computing...ices-next-year

It's only useful, in the same room. It runs at 60GHz and cannot
penetrate walls. So you can't run your whole house off it. But
if you wanted a way to say that "my Wifi is better than my
GbE connection", that's one way to do it. For most people though,
they'll still be using 802.11ac to Wifi up their houses.

Paul

On Sat, 15 Nov 2014 02:29:56 -0700, Chuck Anderson
wrote:

Linea Recta wrote:
I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?

Could be. My wired Ethernet NIC card limits me to a 100MB bandwidth (not
Gigabit). My wireless card, however, will do up to 300MB.

Those numbers would make a LOT more sense if they were written as Mb and not
MB. Mb is accepted to mean megabits per second, while MB is accepted to mean
megabytes per second, a difference of a factor of 8.

--

Char Jackson

On Mon, 17 Nov 2014 21:29:44 -0700, Chuck Anderson
wrote:

VanguardLH wrote:
Chuck Anderson wrote:


Linea Recta wrote:

I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?

Could be. My wired Ethernet NIC card limits me to a 100MB bandwidth (not
Gigabit). My wireless card, however, will do up to 300MB.

My Wireless be faster (vroooom) than wired. ô¿Ô¬


But what bandwidth service tier do you pay from your ISP? Having a
bigger pipe on your end doesn't force their pipe to get bigger. If they
provide only 25 Mbps for what you pay them, you having a 100 Mbps or 1
Gbps pipe won't magically force them to give you any more than the 25
Mbps that you pay for.


D'oh! I didn't let the words "Internet speed test" really register and
was referring only to LAN speeds.

Of course Gigabit or 100Mb makes no difference if the WAN source - ISP -
is well under 100Mb.

Never mind.

I wouldn't say LAN speed "makes no difference". I personally like the fact
that my LAN is mostly Gigabit (while my ISP connection is far less than
that) because it means I can fairly quickly move files around to different
hosts on the LAN.

--

Char Jackson

On Mon, 17 Nov 2014 21:31:21 -0700, Chuck Anderson wrote:

Char Jackson wrote:
On Sat, 15 Nov 2014 02:29:56 -0700, Chuck Anderson
wrote:


Linea Recta wrote:

I was wondering wether the following is normal.
I have a (windows 7) PC connected by cable to the VDSL modem.
I have a (Vista) notebook connected to the modem by WiFi.
Recently I did an internet speed test on both and I saw that the
notebook by WiFi scores faster than the PC.
Something wrong with the PC configuration?

Could be. My wired Ethernet NIC card limits me to a 100MB bandwidth (not
Gigabit). My wireless card, however, will do up to 300MB.


Those numbers would make a LOT more sense if they were written as Mb and not
MB. Mb is accepted to mean megabits per second, while MB is accepted to mean
megabytes per second, a difference of a factor of 8.



Ack., as in ... acknowledged. Ack!

Sorry, I can't resist: as long as we're on the subject of upper vs lower
case, Ack as in Acknowledge should be all caps, ACK :-)

Then it's even unambiguous (or maybe not).

--
Gene E. Bloch (Stumbling Bloch)