Can you give us a little background about yourself?
I suspect I have taken a somewhat unusual career path, starting with a degree in EE and Economics from Darmstadt Institute of Technology (now University of Darmstadt) in Germany. This was followed by graduate degrees in EE from the University of Cincinnati and the University of Massachusetts, Amherst. I had first worked on signal processing, primarily audio compression, and then developed an interest in networking as the Internet was still largely an academic pursuit, well before home Internet connectivity was common. After my PhD and stints at Bell Labs and the German research institute GMD FOKUS, now Fraunhofer, I landed a faculty position at Columbia University in New York. Since 2010, I have also worked in multiple positions at the US Federal Communications Commission (FCC) as Chief Technology Officer and Senior Advisor for Technology. In these positions, I advise Commission staff and leadership on technology issues.
How did you become interested in pursuing a career in Engineering and Technology?
I started out playing with various electronics hobby kits, e.g., by Philips, in high school. I was lucky to end up in a high school exchange program in Livermore, CA, near the Lawrence Livermore Laboratory. I then helped my host uncle assemble a hobby kit 8080 computer in 1978, and had the chance to use a DEC and Cray-1 computer through a 150 baud home modem. I also learned Fortran and BASIC in high school. I then did various electronics projects, such as a printer driver for an old IBM Selectric, before enrolling in an electrical engineering program at TU Darmstadt.
What is it like to be working at the Columbia School of Engineering and Applied Science?
I currently work for two places – the United States Federal Communications Commission (FCC), a US government agency located in Washington, DC, and the Columbia University Department of Computer Science. The two positions are rather different, although the subject matter overlaps. At the FCC, I’ve been working with staff, often economists and lawyers, on a variety of projects in the Internet space, such as Open Internet, next-generation emergency calling (NG911), text-to-911, cybersecurity, relay services for people who are Deaf or hard of hearing, telephone numbering, preventing telephone robocalls and Internet performance measurements. Day-to-day, this includes discussions with other members of staff or attending presentations by outside parties, drafting summaries of technical options for rulemakings, presentations, participating in standards meetings at the Internet Engineering Task Force (IETF), or planning and monitoring contracts for software development for relay services. Often, this involves translating engineering issues so that lawyers and other non-engineers can understand the key issues. On occasion, this may also involve a hearing in front of a committee of the US Congress.
At Columbia University, faculty responsibilities involve teaching, research and various administrative duties. I typically rotate among teaching three courses: Advanced Internet Services (which covers VoIP), Internet Technology, Economics and Policy (which reflects what I learned working at the FCC) and the introductory Computer Networks course, covering all the standard networking topics from applications to Ethernet. I currently focus my research on the Internet of Things (IoT) and video streaming. For IoT, we are looking at how to make such devices interoperable, secure and easier to manage, as part of our SECE (“Sense Everything, Control Everything”) project. For video streaming, we have measured how YouTube stalls (“rebuffering”) affect the behavior of viewers, in a project called YouSlow. Research also involves giving talks at technical conferences, e.g., organized by IEEE, which gets me to travel with some regularity. Unfortunately, research also requires obtaining funds, typically by applying for research grants with various US government agencies such as NSF. (US universities do not fund research directly; almost all research funding is external, either corporate or government.) Most hands-on research is done by graduate research assistants, students that are working towards their doctorate or MS degree, advised by faculty such as myself.
You have a variety of research interests such as Internet real-time and multimedia services and protocols, ubiquitous computing, and mobile systems; can you tell us why you became interested in these topics?
One lead to another – I was initially interested in audio and video coding, as part of my MS thesis work. This was before soundcards for PCs were widely available or usable for real audio capture, so I had to use A/D converters on a PDP-11/44 minicomputer. As the Internet and early Sun workstations such as the SPARC became capable of carrying real-time audio and video streams, I started working on creating and standardizing protocols for carrying audio and video over the Internet, leading to the work on the Real-Time Transport Protocol (RTP) and the Session Initiation Protocol (SIP). The increasing use of mobile devices then led to work on ubiquitous systems, including what has now become known as the Internet of Things. (I had worked on designing software and hardware peripherals for small microcontrollers, such as the Z80 and 8051, early on, so this seemed like a natural evolution.)
You’ve been involved with several research and publications, but you are most well known for your work on Voice Over IP and Session Initiation Protocol, can you tell us about it?
I was working on a PhD research project called DARTnet, a project to carry voice and video over an experimental overlay network running a now-defunct new networking protocol called ST-II. The network nodes were early SUN Microsystems SPARC workstations that implemented network routers and an overlay network. This system needed tools and protocols, and I got interested in working on various aspects of voice and video, from the protocols for transmitting voice or video to algorithms to create smooth playout. I got involved in the Internet Engineering Task Force (IETF), as they were starting to develop the necessary protocols to support the multicast backbone (Mbone), an early way of delivering audio and video to hundreds of viewers. This then led to the development of RTP, initially meant for multicast and then adopted for most other standards-based audio and video delivery. As this work matured, a number of us recognized that we needed better ways to initiate audio and video sessions, leading to the Session Initiation Protocol (SIP). The small group working on this topic had the advantage that nobody cared since “real” telecommunication engineers were working on industrial-strength time-division multiplexing circuit switches. Thus, progress could be relatively rapid, and the underlying standards were ready as the cable and wireless industries started to see the need to move to IP-based networks.
What is it like working at NYU’s Center for Advanced Technology in Telecommunications?
The Center for Advanced Technology in Telecommunications is mostly a “virtual” center, i.e., a collection of faculty and researchers working in the broad area of wireless communication, networking, cybersecurity and adjacent disciplines. This is not a lab or research center as such, but provides a good way to meet and work with other faculty interested in similar topics.
You have received notable awards including the Internet Hall of Fame Innovators in 2013 and the ACM Fellow in 2014. Can you tell us about them?
I have been honored these awards, but they really recognize the efforts of lots of collaborators and students, both at Columbia University and the IETF.
What particular project/research was your favorite?
It’s hard to pick a favorite, but I enjoyed the early work building software tools for network voice (NeVoT), while being able to contribute to standards. More recently, the work on next-generation 911 has been satisfying since it solves a real societal problem.
What part of your job do you find most satisfying? Most challenging?
At my FCC job, the most satisfying part is working on problems that make some small part of the communications infrastructure work a bit better – whether that’s reducing illegal robocalls or supporting the deployment of real-time text for people who are Deaf. The challenging part is the long time scale that such efforts take – I started working on next-generation 9-1-1 (emergency calling) fifteen years ago and deployment is proceeding very slowly. Any advances are also subject to reversal or delay if the political environment changes.
For my university position, working with students to develop new and interesting prototype systems is most rewarding, seeing students gain understanding and skills along the way. Dealing with grading and, in particular, academic dishonesty are easily the most frustrating parts.
What are your favorite hardware and software tools?
I don’t do much hardware development these days, but Wireshark remains my tool of choice to see why network applications are not working as they should.
Do you participate in any professional organizations? Can you tell us about them?
Since my student days, I have been a member of the IEEE, joining the ACM later as my interests started to drift into computer science territory. Together, both probably publish the vast majority of networking-related research. Even though we might consider them publishers and organizers of conferences, they are fulfilling an important social and organizing role, as conveners of engineers and computer scientists with common interests and as a way to propagate, discuss and translate research findings into reality and as a way of knowledge becoming “common knowledge”.
While not a professional organization in the strict sense, the Internet Engineering Task Force (IETF), technically a standards-development organization (SDO), also serves a somewhat similar role, as participants, with the same core returning year after year, meet several times a year and discuss common concerns related to networking standards.
What do you usually do during your free time? Do you have any hobbies?
To keep my hand in doing hands-on (software) engineering, I run a small family business, EDAS Conference ahaServices, which manages engineering and other conference submissions. With my family, I like to spend time outdoors – hiking, kayaking and biking. (I currently enjoy the luxury of getting to commute to work in Washington, DC.)
How about books that you would definitely recommend?
As a non-technical book, T. Wu’s “The Master Switch”, gives a nice summary of communication history, illustrating that problems and developments in the Internet era are echoes of the earlier development of the phone system, movies, radio and TV. For a general introduction to networking, Kurose/Ross, “Computer Networking – A Top-Down Approach” is hard to beat. To see what it took to build a nationwide communication system before the Internet, the 1983 volume “Engineering and Operations in the Bell System” provides a fascinating snapshot.
What do you think is the direction that the Electrical/Electronics Industry is heading towards in the next few years?
Given the diversity of the industry, it’s probably a fool’s errand to make a single prediction across the whole industry, but I suspect three broad trends in the area of networking: (1) commoditization of hardware; (2) common platforms across previously disjoint industries and (3) increasing emphasis on cybersecurity, possibly forced by regulatory and legal incentives. For example, the networking industry is rapidly consolidating and having its margins squeezed since standard hardware platform combined with standard protocols can handle a large fraction of enterprise, data center and carrier needs. It used to be the case that each vertical, whether automotive or building controls, had its own network protocols and related standards, programming languages, operating systems and suppliers. But cost effectiveness, integration across verticals and the need for better software quality on shorter timescales, particularly with a need for much better cybersecurity, will force vendors to fall back on commodity building blocks since nobody will want to pay a premium for a custom operating system with outdated or untested security running on custom hardware running 1990’s-quality standalone software that cannot be readily updated or integrated into larger systems, e.g., as part of data feeds.
Is there anything you’d like to say to the young people to encourage them to pursue a career in Computer Science and Engineering?
It is almost impossible to predict what technology will precisely look like in 20 or 30 years, but it is clear that technology, and its responsible use, will play a key role in solving or at least ameliorating key human problems, whether limiting global climate change, dealing with cyber attacks, traffic accidents or reducing the impact of chronic diseases. Each of us can try to contribute to addressing these problems, even if our individual contribution is likely to be small. A background in computer science and engineering is probably the most versatile preparation even for many non-engineering careers, whether in law, business, medicine, entertainment, sports or teaching.
Engineering and computer science is about working with other people to solve hard and interesting problems. Often, engineering and computer science are seen as solo pursuits, but while focused, individual effort is indeed necessary; much of real engineering outside the classroom is about working with colleagues, customers or students.
- The Way I Work: Interview with David Haboud, Product Marketing Engineer at Altium
- The Way I Work: Interview with Natasha Baker, founder of SnapEDA
- The Way I Work: Interview with Todd Dust, Senior Staff Systems Engineer at Cypress Semiconductor
- The Way I Work: Interview with Bel Haba, an engineer with 400 patents