Interview with Fatemeh Afghah

Featured Engineer

Fatemeh Afghah

Fatemeh Afghah - PhD Student and Research Assistant at ECE Department, University of Maine.

How did you get into electronics/ engineering and when did you start?

My parents always wanted me to be a physician, but as far as I remember I preferred to be an engineer. I remember when I was a kid, I asked my father about TV. I was so curious about how TV works and how we can watch peoples’ picture on TV. My father told me that TV received the signal via antenna, so we can watch movies and listen to the music. And I was like, Signal?! What is that! It did not make any sense to me how movies and music can travel on the air. I wanted to know how antenna works. I was very interested to know how computers and other electrical devices work. To find the answers to all these questions I have decided to be an electrical engineer.

What are you currently working on?

My PhD research is about the applications of Game theory in Relay networks and Cooperative communication systems. Cooperative communications has become one of the most important topics in communications in recent years. In Ad-hoc networks, where different nodes try to send their packets to a central base station, packet collision is unavoidable. At the first glance, it seems that the best option for each user is to send its own packet and to behave selfishly. In this case, they compete to use the finite available resources; so their simultaneous attempts to access the channel produce interference and degrade the system performance in terms of achievable throughput. Hence this is not an optimum choice from network point of view.

However, in cooperative communications users might consider the overall system performance and sacrifice their own preferences. Consequently they cooperate with one another and provide a level of relaying service that improves the total network performance.

One commonly used optimization tool to solve the trade-off problem among users with conflicting goals is Game Theory. Game theory is widely used in different sciences such as economics, politics, biology and engineering. In the game theoretical approach, the problem is modeled as a game where the users act as the players. In this approach the players can make appropriate decisions based on their observations of the game situation including other players’ actions. With the recent advances in digital technology, CPUs embedded in sensor nodes are capable of running the game theory algorithms. To have a better sense of how game theory can be applied in the problem solving, let us consider a party that everybody talks to his/her partner. If each pair behaves selfishly, they would talk loudly to better understand each other. Their conversation appears as an annoying interference to the other pairs and in turn they also start to talk loudly. This cycle continues again and again and at last everybody screams and nobody can hear his/her partner. It is a clear example that the selfish behavior is not usually the best option. There is the same problem in power allocation of sensors in an ad-hoc networks. Basically each sensor prefers to increase its power to have a better signal transmission quality, but on the other hand this will appear as interference to other sensors and degrade the system performance. The optimum solution of this problem can be found by the game theoretical approach. In most of these cases, the optimum power allocation is the Nash Equilibrium solution of the corresponding game.

What are your favorite hardware tools that you use?

My current research is pretty much theoretical and I rarely use special hardware. The only hardware settings I often use are my laptop and papers! For my current research, I usually spend my time on reading textbooks and the recently published papers. I also define different scenarios on drawing board and perform simulations with software tools such as MATLAB and C++.

Before starting my PhD, I was working for a wireless communication company on a GSM network optimization project. That was one of the most interesting projects I have ever done. The project had different parts including BTS site visit, hardware check, antenna direction test, drive test as well as feature test and parameter tuning. We used different hardware in this project including GPS for site positioning, tilt meter for antenna angle tuning, Anritsu SiteMaster for radio link test. But my favorite hardware was TEMS Investigation Tools. It was such as an amazing test tool. With this test suite, we could monitor all the signaling messages at different layers transferred between BTS and cell phone in every detail. We were able to observe different network processes including call setup, handover, location update, and authentication. It was very useful for troubleshooting and analyzing network problems such as call drop and authentication failure. The data collected by this test suite were being analyzed during drive test to solve basic problems and afterwards it was being further analyzed in order to perform network optimization.

What are your favorite software tools that you use?

In mobile network optimization project, I had the chance to work with some of the best wireless network optimization software tools such as TEMS Investigation, ASSET, MAP Info, and NED. But my favorite software was the one we developed in our company to analyze the OMC (Operations and Maintenance Centre) output data. This data was in KPI (Key Performance Indicator) format, which was so complicated, and it was hard to be analyzed. The program processed this data and coverts it to a nice user friendly graph format.

For my current research, I mostly use MATLAB and C++. I love MATLAB because it is really user-friendly and you don’t need to be a professional programmer to work with it.

What is the hardest/trickiest bug you have ever fixed?

It is kind of hard to answer this question and it made me to think about all the projects I have done so far. I could say the biggest responsibility I had was the frequency plan for a mobile network. For those may not be familiar with the interference concept, for instance when neighbors’ antennas work on the same or near frequencies, they will interfere with each other and this will decrease the network performance. During the network expansion, new BTS antennas are added to the network. Their frequencies are selected to minimize the interference with neighbors as much as possible. Otherwise it may cause interference with the existing antennas and degrade the network performance. This project was a totally different experience for me. As a student, all I had done before that was simulation-type projects. In simulation projects if you make a mistake, you won’t blow up anything, but in a real project a small mistake may cause a big problem in the network. For example, wrong settings of a parameter may affect an entire network and disconnect many calls. It was a big responsibility to redesign the frequency plan for a city when with 200000 residents using this mobile network. It took me a week of hard work with lots of stress, but it turned out really good. Being in charge of a city network helped me learn what is really takes to be a communication system engineer.

What is on your bookshelf?

At university, I have some books about wireless communications and game theory such as “Fundamental of Wireless Communications” by David Tse & Pramod Viswanath, “Digital Communications” by John Proakis and “Introduction to Game Theory” by Osborne.

But at home, I have lots of books on history. I love history; I believe we can learn so much from it. I like to read scientists’ biographies; I have many of those. It always keeps me motivated to work hard and never be disappointed. But I should say not all these biographies are happy stories; nothing makes me sad as Evariste Galois’s life story does.

Do you have any tricks up your sleeve?

Well, I am not sure I can call it a trick or not, but to me, when there is a problem in my work that I try to solve it for a day or so, and it makes me so frustrated, I will try to give myself a break and don’t think about that for a while. Usually when I get back to the problem with a refreshed mind I can look at my work from new angles and find a solution very fast. Sometimes there is a small problem which you cannot see when you are tired. So I think it is’ not a good idea to fight with a stubborn problems for days without a break.

What has been your favorite project?

My current research is my favorite project.

What challenges do you foresee in our industry?

As a student, I may not be in the best position to foresee the industry challenges, but based on my experiences, I think there is a gap between what we learn at university and what we are supposed to know as an engineer. In my opinion, hundreds of papers are published every year whose results are not useful in practical applications. This shows that some academic research may not follow the industry’s challenges.