60 Seconds with Professor
Carol Davids

Carol Davids is Director of the VoIP Laboratory at the Illinois Institute of Technology (IIT) in Chicago. She worked in telecom for over 30 years: 22 years at Illinois Bell/Ameritech, where her projects included engineering packet switched data networks and gateways to the Interexchange carriers. For the past decade she has focused on methods for providing telephone services over broadband networks. She has designed and delivered various academic courses in telecom and Internet protocols and architectures, both as an independent consultant for service providers and vendors and as part of her current work as professor of ITM. At IIT’s Center for Professional Development (CPD) she designed and built the VoIP lab, and as Director of VoIP Activities there she is now developing programs and projects linking that lab and its students with industry partners.

Richard “Zippy” Grigonis recently spoke with Prof. Davids about SIP testing, one of the themes of this issue.

RG: For a small lab you’ve dealt with some impressive companies.

CD: I have ongoing relationships with Reef Point, Lucent, Motorola (News - Alert) , and a small company called NESS [Network Expert Software Systems, Inc.], which does surveillance software and makes the monitoring software for AT&T (News - Alert) for their 911 surveillance group. They monitor the network, so their concern is less with the 911 messages than it is with the health of the network. Their software is in my lab because one of my student projects this semester is to expand the capabilities of that software so that it can be applied to an IP network which is doing SIP phone calls.

We have different relationships with different companies depending on their needs and resources. For example, Lucent has a sort of ‘helpmate’ relationship with us. We have inherited from them a lot of infrastructure equipment, much older computing gear and even some of their older VoIP gear. They’ll give us a suggestion, like, ‘Why don’t you develop some SIP application that will be of use to people who were disabled’. So the students built a SIP phone for the hearing impaired. That was a project that took about four semesters. By the time they finished it they had developed and demonstrated code that a hearing-impaired person could use to make a phone call to a pizza delivery service and the pizza delivery guy didn’t have to do anything special to get the call. They ordered the pizza via speech-to-text and text-to-speech translation, which is all moderated by SIP signaling. There was no need for the old TTY [TeleTypewriter] technology.

A company can simply suggest something to us and we’ll go off and develop it. What’s great about that kind of work is that a company sometimes really can’t make the case and come up with a good ROI for dumping a lot of money into certain research, and so they can come to us and the students are delighted to do it. I don’t need to demonstrate ROI, I just need to have a nice piece of code at the end of a semester or two, and a demonstration of it.

RG: In this issue we mention Reef Point and Spirent. You’ve dealt with both companies.

CD: Reef Point came to us and they were interested in third-party independent testing. They brought a device into our lab and their trained our first semester students on how to operate it. The students work in a tight loop with the Reef Point engineers, so that when they find that we’ve crossed some threshold in performance, they’ll be shooting emails immediately to Reef Point’s Scott Poretsky, who in turn finds the developer to fix the code. So, the students really have an experience here of being a test group associated with a product. The kids love it when I give them a real focus and help them innovate.

We also have a Spirent Abacus on permanent loan. It enables us to create fairly large and intense assaults of SIP telephone calls across an application layer gateway.

RG: What’s happening now?

Recently we started working with Motorola. They have little MultiMode, MultiMedia SIP/IMS endpoints. We’re going to do performance testing on them, and we’re setting up our infrastructure in the lab now. In our first semester we’ll build a simple WiFi (News - Alert) network over which we can then observe the performance of these endpoints and then subsequently we’ll probably do some work in Motorola’s labs. First of all, we’ll benchmark just how these things operate, and then subject them to different stress patterns.

Keep in mind that we don’t charge people to do this, because we see ourselves as an education process. We ask for the equipment, frequently, and we’re delighted to get donations to support the lab and its valuable work. Companies always figure out one way or another to reimburse us.