In the summer of 1984 I attended a seminar presented by Bob, a member of technical staff. I carefully listened and realized that my project would be changing some fundamental assumptions related to his talk. When the Q&A time arrived I asked this question– “If Grade of Service (GOS) is no longer a constant, i.e. a variable that changes for different connections, wouldn’t that have an impact on your results?” The speaker acknowledged my question with a brief response; I left the seminar feeling satisfied. After approximately a month into my summer job I understood enough about telecommunication networks to ask an intelligent question. What happen next surprised me.
About a week later, Arik, another member of technical staff, rushed into my office and stated “How did you know?”– referring to the question I asked last week. My question had prompted him to do some analysis taking into the account GOS as a variable. I assume the results had been significant enough to warrant Arik coming to my office. He paid me a complement; Arik spent his time to answer my question because he had become intrigued. So how did I know?
In the summer of 1984 I worked at the Bell Labs West Long Branch Campus. The group I joined consisted of engineers and mathematicians who specialized in telecommunication networks. They had begun to explore dynamic non-hierarchical routing (DNHR) for long distance phone networks. The challenge with any infrastructure consists of balancing the needs of the users versus the expense of the providers. Infrastructure can be a network of roads, sewer pipes or telephone lines. A phone call between Rockville, Maryland and Portland, Oregon would need to travel over a series of trunk lines and routing switches.
Hierarchical arrangement of routing calls had been the norm. Non-hierarchical implied not being restricted to a fixed set of routes. Routing dynamically implied that the routing could change depending upon the traffic demands. This needed to be done without sacrificing quality metrics- Grade of Service being a key one.
My summer job assignment– modify a FORTRAN program that implemented an DNHR algorithm to reflect different Grades of Service. This non-trivial task involved translating the constant GOS into a matrix of values associated with each route. This change involved making other changes to the calculation method of the algorithm. My technical supervisor, Magda, supplied me with background material which included using the Bernoulli distribution to represent network traffic. I figured out the coding changes required, navigated the Unix operating system and compiled the FORTRAN code on a regular basis. This project enabled me to build upon material I had learned studying electrical engineering at the University of Maryland. Specifically, I applied my knowledge communications, statistics and numerical analysis. I enjoyed wrapping my mind around a new way to route traffic. My work enabled me to ask a probing question during the seminar.
Thirty years later I still recall Arik coming to my office with– “How did you know?” I have cherished this moment because it buoyed me early in my engineering career. Every time I doubted my abilities I would think back to that moment and know “I have the right stuff.” Not just to be an engineer but to be a inquisitive even when you’re new to field. While my engineering career took me in a different direction than telecommunications, I continued to ask relevant questions. The engineering fun continued when I sought the answers.
Have a productive day,
Dear Reader, What memory or question does this piece spark in you? Have you had an opportunity to ask a question of an expert? Has someone you admired paid you a complement regarding your technical acumen? Please share your comments or stories below. You too can write for the Engineers’ Daughter- See Contribute for more Information.
Read this to learn more about Bell Labs which is now known as Nokia Bell Labs.
US Patent 4345116 appears to be the first patent that describes Dynamic Non-Hierarchical routing
Many technical papers exist on DNHR. From A T & T Bell Labs Technical Journal here’s a paper comparing hierarchical vs non-hierarchical networks.
In 1996 in IEEE Communications volume 33 Issue 7 Arik co-authored Dynamic Routing in the multiple Carrier International Network, with the following abstract:
Presents an overview of the efforts of the 13 participating carriers that comprise the Worldwide Intelligent Network (WIN) Dynamic Routing Group to introduce dynamic routing internationally. The authors propose that the introduction of dynamic routing could be accomplished in stages, starting with the deployment of flexible overflow routing for bilateral networks. Then the authors examine performance feasibility issues associated with multiple carrier transit routing and analyze data items that need to be exchanged by international carriers to realize dynamic traffic control. They also review major conclusions that can be made based on the results of the first WIN dynamic routing trials. Finally, they review a proposal for standardization of certain signaling messages that enables the realization of real-time state-dependent routing in a multiple carrier environment