Data Engineering

Amir Behrouzi-Far

** ***** ****, **********, **, 08854, USA

Ó +1-732-***-**** Q adelgw@r.postjobfree.com Research Interests

Distributed Systems Queuing Theory Probabilistic Modeling Reinforcement Learning Education

PhD candidate in Electrical & Computer Engineering 2017-present Rutgers University, Piscataway, NJ, USA GPA: 3.9/4.0 Advisor: Prof. Emina Soljanin

Co-Advisor: Prof. Roy Yates

M.Sc. in Electrical & Electronics Engineering 2014-2016 Bilkent University, Ankara, Turkey GPA: 3.94/4.0

Advisor: Prof. Ezhan Karaşan

B.Sc. in Electrical Engineering 2008-2012

Iran University of Science and Technology, Tehran, Iran GPA: 16.48/20 Advisor: Prof. Bahman Abolhassani

Internship Experience

Wireless Information Network Laboratory, North Brunswick, NJ Summer 2017 Project supervisor: Ivan Seskar

– Worked on distributed channel estimation techniques for massive MIMO receiver with multiple computing servers.

Publications

[1] Mehmet Fatih Aktas, Amir Behrouzi-Far, Emina Soljanin, and Philip Whiting. “Load Balancing Performance in Distributed Storage with Regular Balanced Redundancy”. Submitted to IEEE Transactions on Information Theory. 2020.

[2] Amir Behrouzi-Far and Emina Soljanin. “Efficient Replication for Straggler Mitigation in Distributed Com- puting”. Submitted to IEEE/ACM Transactions on Networking. 2020.

[3] Amir Behrouzi-Far, Emina Soljanin, and Roy D. Yates. “Data Freshness in Leader-Based Replicated Storage”. To appear in International Symposium of Information Theory (ISIT) 2020. 2020.

[4] Mehmet Fatih Aktas, Amir Behrouzi-Far, Emina Soljanin, and Philip Whiting. “Load Balancing Performance in Distributed Storage with Regular Balanced Redundancy”. In: arXiv preprint arXiv:1910.05791 (2019).

[5] Amir Behrouzi-Far and Ezhan Karasan. “Dynamic Resource Allocation and Activity Management for Energy Efficiency and Fairness in Heterogeneous Networks”. In: The 6th International Workshop on Smart Wireless Communications (SmartCom 2019).

[6] Amir Behrouzi-Far and Saeideh Mohammadkhani. “Cooperative Beamforming in Cognitive Radio Relay Networks Using Amplify-and-Forward Relaying Technique”. In: The 6th International Workshop on Smart Wireless Communications (SmartCom 2019).

[7] Amir Behrouzi-Far and Emina Soljanin. “Data Replication for Reducing Computing Time in Distributed Systems with Straggles”. In: Accepted in IEEE Big Data (2019).

[8] Amir Behrouzi-Far and Emina Soljanin. “Redundancy Scheduling in Systems with Bi-Modal Job Service Time Distributions”. In: 2019 57th Annual Allerton Conference on Communication, Control, and Computing (Allerton). IEEE. 2019, pp. 9–16.

[9] Amir Behrouzi-Far and Emina Soljanin. “Scheduling in the Presence of Data Intensive Compute Jobs”. In: Accepted in IEEE Big Data (2019).

[10] Amir Behrouzi-Far and Doron Zeilberger. “On the Average Maximal Number of Balls in a Bin Resulting from Throwing r Balls into n Bins T times”. In: arXiv preprint arXiv:1905.07827 (2019).

[11] Amir Behrouzi-Far and Emina Soljanin. “On the Effect of Task-to-Worker Assignment in Distributed Com- puting Systems with Stragglers”. In: 2018 56th Annual Allerton Conference on Communication, Control, and Computing (Allerton). IEEE. 2018, pp. 560–566.

[12] Amir Behrouzi Far and Abolfazl Falahati. “Maximizing outage capacity in rayleigh fading channel”. In: International Journal of Computer Science and Artificial Intelligence 2.1 (2012).

[13] Amir Behrouzi Far, Alireza Mohammad Shahri, Saeed Karimi, Hamideh Esmaeili, Majid Haji Mohammadi, and Sepideh Hosseyni. “A Large Covered and Simple To Implement Local Positioning System”. In: 1st joint conference of Robotics and AI and the 3rd Robocup Iran Open International Symposium. 2011. Key Research Projects

On the policy update rate in reinforcement learning Spring 2020-present Advisor: Prof. Emina Soljanin and Prof. Roy Yates

– Implemented a reinforcement learning framework in Tensorflow.

– In progress.

Characterizing Data Freshness in Leader-Based Replicated Storage Fall 2019-present Advisor: Prof. Emina Soljanin and Prof. Roy Yates

– Modeling the write and read operations in leader-based replicated storage.

– With age of information as the metric, the average staleness of the retrieved data by a read query was analyzed. Scheduling in Queuing Systems with Bi-Modal Job Service Time Distribution Spring 2019-present Advisor: Prof. Emina Soljanin

– Effect of redundancy scheduling on the delay performance in queuing systems were studied.

– Using classical Urns&Balls problem performance indicators for scheduling policies were introduced.

– A new scheduling policy, based on combinatorial block designs was introduced.

– With simulation, it was shown that the proposed policy considerably outperforms the existing. Computing Time Analysis of Distributed Computing Systems with Stragglers Spring 2018-Spring 2020 Advisor: Prof. Emina Soljanin

– Timing performance of distributed computing systems, in the presence of stragglers, was studied.

– For several service time distributions the optimum redundant task assignment was derived.

– The variation/predictability of performance was analysed and intuitions for robust system design were provided.

– For different performance metrics optimum redundancy level was derived.

– Significant improvements were observed over Google cluster traces by applying optimum redundancy schemes. Energy Efficiency and Fairness Improvement in Wireless Heterogeneous Networks Fall 2015-Fall 2016 Advisor: Prof. Ezhan Karaşan

– The cell-edge deployment of small cells in cellular heterogeneous systems were studied.

– Dynamic activation of small cells, for improving the energy efficiency of network, was studied.

– Optimization problem for maximizing cell-edge utility and energy efficiency was formulated.

– The optimization problem was implemented in Matlab and solved using Baron software. Selected Graduate Courses

Stochastic Processes Data Mining and Machine Learning Convex Optimization Information Theory Detection and Estimation Theory Linear Systems Theory Communication Networks Digital Communications Error Control Coding Selected Taught Courses

Probability and Random Processes Network Centric Programming Wireless Communications Programming Skills

Fluent in Python: Tensorflow, SciKit-learn, PySpark, Pandas, PySpark, NumPy, SciPy, SimPy, MPI4Py, threading, socket, MatPlotLib.

Experienced with SQL, C, C++, R and Matlab.

Contact this candidate