Your real answer is: It depends.
There's a lot of good information here, but it is mostly skirting the issue, which is whether an IP can be traced to you. The issue really is HOW
CLOSELY can an IP address be traced to you, and this depends on how IP addresses are allocated, either dynamically or statically.
STATIC IP ADDRESSES
The networks are classified as Class A, Class B, or Class C. There are only a few large Class A networks. These can be subdivided into what amounts to
Class B and Class C networks, and Class B networks can be subdivided into Class C networks. Many large corporation have entire Class B networks
assigned to them. Many smaller organizations have one or more Class C networks assigned to them. IN THE OLD DAYS networks were assigned in a liberal
manner.
IN THE OLD DAYS I was in charge of IT at a library system that had ten branch libraries. They ranged in size from about 1,000 square feet with five
"IP-addressable devices" (known as computers) to 35,000 square feet with less than a hundred "IP addressable devices," known as computers. I asked for
and was awarded one Class C Network for every branch, plus an additional one to tie these ten Local Area Networks (LANs) into a Wide Area Network
(WAN). This was complete and utter overkill, as you will see.
A Class C Network is comprised of 256 numbers. For example, one of my Class C Networks was:
198.187.135.X where the numbers (X) ranged from 198.187.135.0 to 198.187.135.255. Now, you can't use "0" as an address and you can't use "255" as an
address, so that left 254 numbers: 198.187.135.1 to 198.187.135.254 for assignment to my "IP addressable devices," known as computers. So I had 254
numbers to assign, in some cases, to a whopping five computers, leaving 249 free numbers, which, at this point were WASTED numbers.
In my largest branch I had more devices so I made up a scheme. The ".1" ("dot-one") address was reserved for my router. The "dot-tens" were reserved
for my hubs. The "dot-20's" were reserved for my servers, the "dot-40's to "dot-90s" for my staff computers, and the "dot-100's" and above for my
public computers. I just made this up and applied it to every branch. "dot-one" was always my router, etc.
So IN THIS CASE I assigned STATIC IP addresses to a specific computer, and mine was "dot-34," therefore if you found the number 198.187.135.34, you
could pin it to me, personally, and the answer to your question is effectively,
"Yes, an IP address CAN be traced to a person."
DYNAMIC IP ADDRESSES
Now obviously in several years I found myself "resource rich." Those IP addresses were running out. A lot of the reason for this was the slop in the
system as I have described. I had about 100 devices overall and about 3,000 IP addresses. They were MINE! too. No one could take them away from me and
I wasn't about to give them up. Indeed, that would have been foolish, because now I had several hundred "IP addressable devices" including printers,
fax machines, printer servers themselves, etc. And I had about 40 servers doing stuff like payroll, accounting, firewall"ing", virus protection,
inventory control, online public catalog, CD-ROM servers, and suddenly those 254 numbers weren't enough for my largest branches.
So I stole some of the numbers from the very small branches and began to re-number my network to allow me more room to maneuver using a technique
called "sub-masking." But if I had gone back to my "issuing authority" and started over and said, "I want 11 Class C networks they would have said,
"Get a life. You ain't getting them." Indeed, when I had need for some different IPs on a separate network I asked US West (Quest, then Centurylink)
and they gave me 4--reluctantly. So the era of freely-given and wasteful IP allocation was over.
So what is an organization like me to do? Implement DYNAMIC IP addressing. In this case, I would allocate IP addresses to my devices just like before,
but on a temporary basis that had an "expiration." They could use this IP address for a certain period of time, but at some point would have to
"renew" their address, at which point it might not be the same one they had prior. FURTHER, these addresses were for INTERNAL use only. By the time
the signals got to my router, they appeared as if they were coming from a single IP address before they went out into the world, thus a hundred
internal devices could share what amounted to a single IP address. This is now considered the "proper" thing to do, ethically.
This is known as "dynamic addressing" because it dynamically changes, and in this case there is no way to trace the IP address to you, personally, so
your answer is,
"No, an IP address CANNOT be traced to a person."
BUT (there always is one) it could be traced to my library, so if the Bad Guys(tm) were interested in WHERE a given message originated, they would be
knocking at my door knowing that from somewhere inside my network the message was sent. Could I then find out who sent that message? Perhaps. I can't
tell you for certain. We'd have to go through the motions, examine exactly how the network was set up, and try to trace it.
THIS is how an Internet Service Provider allocates IP addresses. They do it dynamically. And it's also why you think you are insulated when your IP is
reported a few cities away from your real location. When I was with Quest my location was usually in a city about 50 miles away called Tukwila. That's
because that's where the Quest switching equipment was located. But if the Bad Guys(tm) were looking for me, they would know that I was within the
Seattle metropolitan area. That's a couple of million people, but it's not 360 million, and at this point they could use other clues to narrow their
search further. And, frankly, who knows what kind of tracing equipment and databases are kept by your ISP?
So, you see that the answer to your question is not a simple one. There are many factors involved, some of them even historical in nature. And that's
why the answer is: It depends.
edit on 3/8/2014 by schuyler because: (no reason given)