I discovered a patent application (not yet granted) this week that gave me chills. Part of its claims* are performing Layer 2 and Layer 3 network topology discovery and display, as of October 2001. Huh? Network topology discovery and display was being done way before then. What's chilling about this patent application? Having a patent granted that covers basic network discovery. That's the foundation of many/all network management products, prior to the filing date and subsequent to then.
Application 10/270122, titled "Network traffic generation and monitoring systems and methods for their use in testing frameworks for determining suitability of a network for target applications", with the provisional application filed Oct 15, 2001.**
Companies and products like HPOV, Lumeta, Spectrum, CiscoWorks, and many others did network discovery and topology display many years prior to Oct 15, 2001. Such pre-existing network topology discovery and display is called "Prior Art" in patent office speak. In fact, the application includes references to many of these tools, yet it doesn't explain (at least to me) how the application's processes are unique, novel, and unobvious to one "skilled in the art" (more patent office speak) -- i.e. patentable. At one point, the application references "Mapping the Internet", IEEE Computer, vol. 32, pp 97-98, April 1999, by H. Burch and B. Cheswick, but then says that it was about mapping the internet backbone and didn't apply to enterprises. How's that? How are routers, switches, and network topology different in an enterprise than in the Internet backbone for the purposes of this patent?
But what about Layer 2 mapping? CiscoWorks incorporated something to do this, but I've not yet found a date (please leave a comment here or on our corporate website comment form if you know of a date). An associate, Kennedy Clark, developed a tool prior to Oct 2001 that used SNMP to query the CDP MIB to learn network topology (both L2 and L3). It is an easy jump for someone skilled in the art to suggest switching to different tables - the core mechanism is the same. The algorithm for doing the topology discovery was certainly in existance at that time - we did it manually for the most part.
While the claims on network and topology discovery are probably unenforcable, it is better to show prior art before it becomes a patent. If you know of prior art (specifiy products or documentation and their dates), please let me know and I'll pass it along to the examiner.
-Terry
* Note: I'm not a patent lawyer. I'm just a technology person reading a patent and determining what I think is applicability. However, I'm representative of the type of expert witness to be called to testify in a case to prove or disprove prior art. I'm not saying that the entire application is invalid; only that the network discovery and topology display claims are invalid.
** I don't know the authors of the patent and have nothing against them. In fact, the primary author looks to be a pretty bright person and has done some neat work on a broad variety of topics.