Slow WiFi speed on Droid Eris 1.5

[youngdoc]

My Eris does not seem to be taking advantage of the WiFi speed capabilities of my home network. I just tested my network in multiple ways. I'm getting good connection speeds on all devices EXCEPT my Eris. Help is much appreciated.

Here's what I did:

1. I went to Speakeasy.net and tested the hard wired internet connection speed to my PC. The result was 22 mbps down.

2. I tested the wireless connection speed to my laptop, also via Speakeasy.net. Again, 22 mbps.

3. I tested the wireless connection speed to my wife's iMac, also via Speakeasy.net. Once again, 22 mbps

4. I tested the connection speed on my Eris (repeatedly) via a speed test app that I downloaded from the Market. Only 2.5 mbps.

5. I loaded the Speakeasy.net website in my Eris' Android browser and ran the same test. Only 2.9 mbps.

So, all wired and wireless devices in the house are getting ~ 22 mbps, while my Eris is getting significantly less.

Specifications:

-Eris is running default Android 1.5 OS

- Home wireless router is a Netgear Wireless-N Router model # WNR2000 running in WPA-PSK [AES] security mode. "Mode" is set to the highest -- "Up to 300 mbps."

What gives?

Many thanks.

 

[doogald]

I just use the speedtest.net app. My ISP has 15 mbps down, 1 mbps up. The Eris is reporting 5.6 mbps up, 684 kbps up. Obviously, the Eris's networking (or CPU) can't handle full speeds, but that's good enough for me.

 

[youngdoc]

Suggestions to correct the problem (as opposed to "WiFi works fine for me") would be much appreciated.

 

[Podivin]

I ran a few tests at the Speakeasy site, using different servers there. I'm seeing ~3mbps down and ~.3 up, which seems about right to me (in fact it seems pretty good). I've not done any research on the matter, but I don't believe it's reasonable to expect the Eris to have the same wifi speeds as a full fledged computer. I suspect there's something buried deep down in a spec sheet somewhere that shows what the max wifi speed is, and I bet it's nowhere near computer speeds.

 

[doogald]

I'd try a couple of things:

1. Get closer to the access point. Perhaps you are too far away to get a strong signal. The further away that you go, the slower the connection speed you will get.

2. As a test, try setting the Netgear to 802.11 g/b and remove 802.11n from the equation. The Eris supports 802.11 b/g only; it does not have an 802.11n radio. You can always set it back to n/g/b mixed mode later.

I meant to do this with my previous post, but just as an FYI, I have three routers acting as access points in the house, all with the same SSID on different channels, all set to 802.11 n/g/b mixed mode, all with WPA2 (so AES encryption.) Two are Apple Airports, one is a Linksys 160N. Because they are all the same SSID, it's hard to tell which one I was connected to, but when I ran that test I was about 15 feet away from the nearest AP and maybe 30 feet from the furthest.

Sorry I can't help more than that. I can think of any number of reasons why throughput would be limited. The CPU is only 528 MHz; there is only 512 MB of RAM. Perhaps the device is giving lower priority to WiFi networking to give priority to phone communications. Perhaps the WiFi radio is poor, or the driver is poorly implemented. That said, you can try the things that I listed above, especially the b/g mode restriction, to see if it helps (and perhaps does not hurt your other devices.)

 

[erisuser1]

youngdoc,

I think you are going to be hard-pressed to find anyone with an Eris that reports something more than 4 Mbits/sec in sustained download rate. (This assumes that they correctly avoid downloading uncompressed files in their benchmarks).

I think that what you are running into is a disconnect between your expectations and experience - so, at some point it is important to ask what your expectations are based upon.

The VZW/HTC specs for the Eris only say: "802.11 b/g".

What they don't say, is "capable of sustained data transfer of 54 Mbits/sec end-to-end including multiple protocol layer translations, application coding, and storage."

I've worked my whole career in microelectronics and system electronics, and I can tell you for a fact that when it comes to publishing specifications, the marketing intent is to create the most favorable view of the product possible... without ever lying outright. If that is done in such a way that the customer mistakenly believes the product is better in some regard than it actually is - well, that's the customer's mistake. I'm not saying that is right or wrong - it's just the way the world works.

When reading specs, a good rule of thumb is the rubric: "if they left it out, you should remain in doubt".

Now, I didn't want to just talk through my hat, so I ran some tests early this morning on your behalf: I set up both a Web Server (Apache) and an FTP server (FileZilla) on my LAN and did some download testing. (HTTP using Eris native browser, and FTP via ES File Manager App; the server was hard-wired to the LAN so the only wireless traffic would be between the AP and the Eris). That also eliminates the wildcard of internet congestion. Further, I set up my 802.11g router to use only "802.11 G" behavior, and fixed transmission speeds.

The result? The Eris is capable of operating at 54 Mbps (AP Tx)... in terms of radio bit rate, anyway. It's sustained transfer rate was only about 3 Mbps, however. Changing the (fixed) AP 802.11g transmit speed to several different rates, all the way down to 6 Mbps, produced only modest changes in transfer rate, with the best sustained data rate of about 3.7 Mbps (for a 50 MB file). As a sanity check, a separate wireless client was used, and a ethernet-to-ethernet check was performed with the same server setups. With my half-decade old equipment, the best it could do was about 17 Mbps (wired), and about half that wirelessly. I didn't carefully check the detailed settings of the other wireless client, so that check only verifies that my half-decade old router/AP can handle more than what the Eris can achieve.***

Anyway - technically speaking - VZW and HTC are well within their rights to say that their equipment is "802.11 b/g" - it in fact it does seem to operate at a radio bit rate of 54 Mbps (TX) by the AP.

I'm sorry if you feel cheated - but neither VZW or HTC made any claims that should make you feel that way. The Eris is running on a chipset designed to use no more than about 2 watts of power in WiFi-only mode (more if the cellular radio is in use)- and you are comparing it to other devices (laptops, desktops) that employ CPUs that dissipate anywhere from 40-150 watts of power. Not a fair comparison, IMO.


eu1


*** These tests were performed using WPA-Personal (TKIP) encryption. It is possible that encryption overhead could be a performance-limiting factor (as are many other factors), but if that was the case, the alternative would be using WEP or no encryption at all - and I wouldn't recommend using either of those strategies to improve your throughput.

 

[dirtmouth]

youngdoc,

I think you are going to be hard-pressed to find anyone with an Eris that reports something more than 4 Mbits/sec in sustained download rate. (This assumes that they correctly avoid downloading uncompressed files in their benchmarks).

I think that what you are running into is a disconnect between your expectations and experience - so, at some point it is important to ask what your expectations are based upon.

The VZW/HTC specs for the Eris only say: "802.11 b/g".

What they don't say, is "capable of sustained data transfer of 54 Mbits/sec end-to-end including multiple protocol layer translations, application coding, and storage."

I've worked my whole career in microelectronics and system electronics, and I can tell you for a fact that when it comes to publishing specifications, the marketing intent is to create the most favorable view of the product possible... without ever lying outright. If that is done in such a way that the customer mistakenly believes the product is better in some regard than it actually is - well, that's the customer's mistake. I'm not saying that is right or wrong - it's just the way the world works.

When reading specs, a good rule of thumb is the rubric: "if they left it out, you should remain in doubt".

Now, I didn't want to just talk through my hat, so I ran some tests early this morning on your behalf: I set up both a Web Server (Apache) and an FTP server (FileZilla) on my LAN and did some download testing. (HTTP using Eris native browser, and FTP via ES File Manager App; the server was hard-wired to the LAN so the only wireless traffic would be between the AP and the Eris). That also eliminates the wildcard of internet congestion. Further, I set up my 802.11g router to use only "802.11 G" behavior, and fixed transmission speeds.

The result? The Eris is capable of operating at 54 Mbps (AP Tx)... in terms of radio bit rate, anyway. It's sustained transfer rate was only about 3 Mbps, however. Changing the (fixed) AP 802.11g transmit speed to several different rates, all the way down to 6 Mbps, produced only modest changes in transfer rate, with the best sustained data rate of about 3.7 Mbps (for a 50 MB file). As a sanity check, a separate wireless client was used, and a ethernet-to-ethernet check was performed with the same server setups. With my half-decade old equipment, the best it could do was about 17 Mbps (wired), and about half that wirelessly. I didn't carefully check the detailed settings of the other wireless client, so that check only verifies that my half-decade old router/AP can handle more than what the Eris can achieve.***

Anyway - technically speaking - VZW and HTC are well within their rights to say that their equipment is "802.11 b/g" - it in fact it does seem to operate at a radio bit rate of 54 Mbps (TX) by the AP.

I'm sorry if you feel cheated - but neither VZW or HTC made any claims that should make you feel that way. The Eris is running on a chipset designed to use no more than about 2 watts of power in WiFi-only mode (more if the cellular radio is in use)- and you are comparing it to other devices (laptops, desktops) that employ CPUs that dissipate anywhere from 40-150 watts of power. Not a fair comparison, IMO.


eu1


*** These tests were performed using WPA-Personal (TKIP) encryption. It is possible that encryption overhead could be a performance-limiting factor (as are many other factors), but if that was the case, the alternative would be using WEP or no encryption at all - and I wouldn't recommend using either of those strategies to improve your throughput.

^^^ that was cool.

 

[pushdrops]

erisuser1, thanks for your detailed results, I've been meaning to do tje same as just like youngdoc, I've been disappointed in the Wi-Fi performance of the Droid and Eris.

i can see why youngdoc feels cheated, because other wireless devices are providing faster throughput speeds than his Eris... , no matter the reason, it's just something most would expect....

Sent from my Droid using ....