I've had two of these routers for a couple of months now (main and vacation home), and the router at the vacation home just started exhibiting this behavior about two days ago.
Both are: DIR-655 (H/W Ver.: B1 F/W Ver. 2.00NA)
Before this both ran without error for almost two months, and it has so many more features than my old LinkSys router (up to 24 dhcp IP address reservation, and port-forwarding entries) that I can't go back to my old LinkSys.
Symptom: The wired side stays connected just fine, but the wireless side goes down about every 3-4 hours. My work-around has been to either do a power reset, or to do a reset through the admin screens on the remaining wired connection. Luckily, the problem router can be controlled by my home automation system so I can do a powercycle remotely when I can no longer ping one of my wireless cameras.
On the problem router, a couple of things to note:
- happened only when I had over 20 different wireless endpoints connected -- everything from PCs, to Androids, to iPhones to iPads to iTouches to colorNook to wireless cameras.
- when I check the log through the admin web interface, it is empty (or at least nothing shows)
- when the wifi stops broadcasting, everything else looks normal -- such as the front panel lights and what I can see through the web admin interface (except it shows no wireless clients and the log is empty)
- now that most of the wireless devices have left the area, it has been running for over 12 hours. So, it could be correlated to one of the devices that was connecting to it. Some of the PCs may have been running various firewalls and the various devices like iPhones, Androids, and Nooks tend to connect/disconnect frequently -- so I don't know if there is some ecosystem cause/effect.
When I compare the config.bin files for both routers. below I note the differences -- I won't be able to do play with this as I won't be back to my problem router for a couple of weeks:
hostname=DIR-655 (this is just the first line in the config.bin file)
...
dhcpd_netbios_enable=1 (this is set to 0 on the router without the issue)
dhcpd_netbios_learn=1 (this is set to 0 on the router without the issue)
...
wan_specify_dns=1 (this is set to 0 on the router without the issue)
wan_primary_dns=204.194.232.200 (this is set to 0.0.0.0 on the router without the issue)
wan_secondary_dns=204.194.234.200 (this is set to 0.0.0.0 on the router without the issue)
...
wish_engine_enabled=1 (this is set to 0 on the router without the issue)
....
anti_spoof_check=1 (this is set to 0 on the router without the issue)
...
wlan0_dot11_mode=11gn (this is set to 11bgn on the router without the issue)
...
wlan0_11n_protection=20 (this is set to auto on the router without the issue)
...
opendns_enable=1 (this is set to 0 on the router without the issue)
...
Since the problem router was running fine with the same settings as shown above for 2 months (only difference is that I added several additional dhcpd_reserve and port_forward_both entries), I am not convinced it has anything to do with the settings.
Could it be one of the endpoints? Many of the same wireless endpoints were also running at my main home (but not all at the same time) before they created a perfect storm at my second home and the wifi started to have issues.
So, my bet is that either my router went bad, or there is something about some of the endpoints that caused the router to have issues. If the wifi continues to run fine without all the endpoints present then I can probably rule out the bad router -- but only time will tell on this one.
If this is the case, then perhaps one of the endpoints caused the wireless part of the router to misbehave ... somewhere in this forum today I found another thread that indicted some firewall software causing a similar issue.