ADSL issue or network issue, we are not sure how to narrow it down further

We have an issue on our ADSL connection, networked to a few (4, sometimes 5 in use at a time) computers, as follows:

* 1 x Linksys Modem Router - Wireless-N ADSL2+ Modem Router WAG120N
* 1 x 24 port switch
* 5 computers connected to the switch via LAN CAT5E cables
* Another few workstations have wired connections also, but are not in use at the moment (i.e. the wires are there but no PCs).
* The wireless of the linksys router is currently NOT in use by anyone. Wireless is still "on" though, unused.

The issue is that our connection keeps dropping. Our computers (mostly on Windows 7, and one on Ubuntu) lose network access, although we still can connect to the Linksys interface 192.168.1.1. The status of the router says "Disconnected" at this time, which means it got disconnected from the internet. The router automatically reconnects after a minute or so. I have pasted below a typical log from the router just before and after the disconnection/reconnection. I have labelled there where the disconnection occured (see below).

We also have an older single port modem-router from D-Link, which does not drop the connections, but has its own problems, such as the internet becomes really slow sometimes. We have called our service provider support who tell us that the phone line up to our premises is fine. They did not check the inside connections though, they just guarantee that the wire uptil the premises is OK. So, we think it's one of the following issues: i) The phone line from the outside to our modem inside (i.e. insider our premises) is faulty, so we need to put in a new/fresh phone line and try this. Noise could sometimes disrupt the connection of course, and one of our guesses is that noise is causing the connection to become slow (old router) or disconnected (new router).

ii) Some issue within our network is another possible cause - on our PCs. We run Avira antivirus and Lavasoft Awaware and Spybot to clean up viruses / spyware etc. and so far Spybot just finds a few browser cookie issues when we run it and cleans them up.
We're lost on how to know exactly what is wrong. Running a tracert on www.google.com works fine when the connection is up, and doesn't work when the connection is down. For example:

C:\Users\THE>tracert www.google.com Unable to resolve target system name www.google.com.

C:\Users\THE>tracert 192.168.1.1

Tracing route to 192.168.1.1 over a maximum of 30 hops

1 <1 ms <1 ms <1 ms 192.168.1.1
Trace complete.

And when the modem router is connected, the tracert has about 11 hopes. A couple of those are say 106ms and 754ms towards the end, I think because generally our net connections here in Sri Lanka aren't that good! I'm on a 1Mbps line and sometimes get a good download rate, but then again we do have the above issues during office hours. Any help here would be appreciated, I'd like as much feedback as possible before attempting to put in a new phone line to see if that is the issue.

The log from the router:

Fri, 2011-09-09 14:20:34 - Returning UPnPError 714: NoSuchEntryInArray

Fri, 2011-09-09 14:20:34 - AddPortMapping: external NULL:42193 to 192.168.1.30:42193 protocol TCP for: MSNMSGR with timeout:0

Fri, 2011-09-09 14:20:34 - no permission rule matched : accept by default (n_perms=0)

Fri, 2011-09-09 14:20:34 - redirecting port 42193 to 192.168.1.30:42193 protocol TCP for: MSNMSGR

Fri, 2011-09-09 14:20:34 - creating pass rule to 192.168.1.30:42193 protocol TCP for: MSNMSGR

Fri, 2011-09-09 14:20:34 - GetSpecificPortMappingEntry: rhost='NULL' 42193 TCP found => 192.168.1.30:42193 desc='MSNMSGR'

Fri, 2011-09-09 14:20:34 - AddPortMapping: external NULL:42193 to 192.168.1.30:42193 protocol TCP for: MSNMSGR with timeout:0

Fri, 2011-09-09 14:20:34 - removing redirect rule port 42193 TCP

Fri, 2011-09-09 14:20:34 - Trying to delete rules at index 0

Fri, 2011-09-09 14:20:34 - DeletePortMapping: external port: 42193, protocol: TCP

Fri, 2011-09-09 14:20:34 - removing redirect rule port 42193 TCP

Fri, 2011-09-09 14:31:26 - No response to 3 echo-requests

Fri, 2011-09-09 14:31:26 - Serial link appears to be disconnected.

Fri, 2011-09-09 14:31:26 - Enter: tdb_store.

Fri, 2011-09-09 14:31:26 - tdb_store: calling tdb_update.

Fri, 2011-09-09 14:31:26 - Enter: tdb_update.

Fri, 2011-09-09 14:31:26 - Enter: tdb_store.

Fri, 2011-09-09 14:31:26 - tdb_store: calling tdb_update.

Fri, 2011-09-09 14:31:26 - Enter: tdb_update.

Fri, 2011-09-09 14:31:26 - Enter: tdb_store.

Fri, 2011-09-09 14:31:26 - tdb_store: calling tdb_update.

Fri, 2011-09-09 14:31:26 - Enter: tdb_update.

Fri, 2011-09-09 14:31:26 - Couldn't increase MTU to 1500.

Fri, 2011-09-09 14:31:26 - Couldn't increase MRU to 1500

Fri, 2011-09-09 14:31:26 - LCP down.

Fri, 2011-09-09 14:31:29 - Failed to get ip address for interface ppp0

Fri, 2011-09-09 14:31:29 - Failed to get ip address for interface ppp0

Fri, 2011-09-09 14:31:32 - Connection terminated.

Fri, 2011-09-09 14:31:32 - Connect time 12.1 minutes.

Fri, 2011-09-09 14:31:32 - Sent 626579 bytes, received 2853159 bytes.

Fri, 2011-09-09 14:31:32 - Enter: tdb_store.

Fri, 2011-09-09 14:31:32 - tdb_store: calling tdb_update.

Fri, 2011-09-09 14:31:32 - Enter: tdb_update.

Fri, 2011-09-09 14:31:32 - Doing disconnect

Fri, 2011-09-09 14:31:32 - Exit. ******** [We lost connection here]

Fri, 2011-09-09 14:32:49 - Adsl is connected ******** [We regained connection here]

Fri, 2011-09-09 14:32:55 - Initialize LCP.

Fri, 2011-09-09 14:32:55 - Plugin pppoe loaded.

Fri, 2011-09-09 14:32:55 - PPPoE Plugin Initialized

Fri, 2011-09-09 14:32:55 - Plugin pppoe called.

Many thanks!

I didn't read the whole log there but I did read what you wrote. One suggestion I could have is that you tried "google.com" and got no response, but tried "192.168.1.1" and did. To me that suggests taking a look at DNS.

Before you have another outage do a "nslookup" for google (and a few others.) keep a short list of some of those IP addresses you get. When you are in your outage period try pinging the name first (you should get an "address unresolved" message) and then try its corresponding IP address. If the IP works and the name does not you have a DNS server issue.