Hossein Mehri

I hold a PhD in Communication Systems from Boise State University's Electrical and Computer Engineering Department, where my research focuses on machine learning algorithms for time-series prediction in next-generation wireless networks. Throughout my academic journey, I acquired a deep understanding of machine learning concepts and mathematics, particularly in time-series prediction and image processing, as well as expertise in Software Defined Networking (SDN), Network Function Virtualization (NFV), and Network Coding through coursework and projects. In addition to my academic pursuits, I have practical experience as a Back-Office Engineer in the telecom industry, where I managed maintenance tasks for Huawei and Ericsson nodes. This role honed my technical skills in script writing, debugging, problem solving, troubleshooting, optimization, programming, and management, while also fostering soft skills such as teamwork, knowledge sharing, stress management, and time management.

• 4G Network Fundamentals, Fall 2023, Institut Mines-Télécom, USA
• Linux Fundamentals, Fall 2023, LearnQuest, USA
• Computing in Communication Networks from Theory to Applications, Fall 2022, IEEE ComSoc Training, USA, 761527
• Java Programming, Fall 2017, Kahkeshan IT College, Tehran, Iran
• CCNA, Winter 2017, NAK World Class Telecom Managed Services, Tehran, Iran
• Network+, Spring 2017, Kahkeshan IT College, Tehran, Iran
• 3G Technical, Winter 2016, NAK World Class Telecom Managed Services, Tehran, Iran
• GSM Basics, Winter 2016, NAK World Class Telecom Managed Services, Tehran, Iran
• MATLAB, Summer 2010, Iran Technical and Vocational University (Zanjan Branch), Zanjan, Iran

• Machine Learning
• Stochastic Signals and Systems
• Mobile Communication I
• Stochastic Processes
• Information Theory and Coding
• Advanced Computer Networks
• Optimization Theory and Practice
• Linear Systems
• Applied Electromagnetics
• Wireless Sensor Networks
• Advanced Telecommunications Theory
• Computer Networks in Telecommunications
• Simulating Telecommunications Systems

• ECE 210 - Introduction to Electric Circuits (Teaching)
• ECE 350 - Signals and Systems (Teaching Assistance)
• ECE 300 - Electromagnetic and Wave Theory (Teaching Assistance)
• Electrical Circuits II (Teaching)
• Mobile Communications (Teaching Assistance)
• Signal and Systems (Teaching Assistance)

• H.Mehri, H.Mehrpouyan, H. Chen, "RACH Traffic Prediction in Massive Machine Type Communications," submitted to IEEE Trans. Machine Learning Comm. Net.
• H.Mehri, H.Mehrpouyan, H. Chen, "Cellular Traffic Prediction Using Online Prediction Algorithms," arXiv, 2024.
• H.Mehri, H.Mehrpouyan, H. Chen, "Asymmetric Autoencoder-based Data Compression for Traffic Prediction," ready to submit.

• Traffic Pattern Prediction Using Machine Learning Algorithms:

      Using LSTM, GRU, Echo State networks to predict the traffic of massive machine type communications. In this project, a new algorithm is proposed in the prediction phase that can correct the machine learning network states using the frequently collected data to increase the accuracy of the predictions without imposing a significant processing load. 

• Developing a High-Performance Data Compression Scheme for 5G Cellular Networks:

      Developed a novel distributed-centralized compression and decompression framework using asymmetric autoencoders to reduce data collection overhead in cellular networks. The scheme outperforms existing methods in processing complexity and reconstruction accuracy. 

• Channel State Prediction in MIMO Transmissions Using Machine Learning Algorithms:

      Applying CNN-LSTM and ConvLSTM models to channel state prediction problem in MIMO transmissions, where a series of complex channel state matrices is considered as series of images (video frames). In this project, the PyTorch models are modified to work with complex numbers. 

• Research on Software Defined Networks and Network Function Virtualization (SDN/NFV):

      Conducting a comprehensive research on the virtualization concept in SDN/NFV networks, as well as investigating the applications of machine learning algorithms in these networks with centralized controller. 

• Designing Webpage Using Java Frameworks and MySQL Data Server:

      Designing a webpage using Java frameworks and MySQL data server, with the ability to add new users with different levels of access. Each user may be able to view, upload, and manage only his or her own files (images, ...) or the files of other users, depending on their level of access. 

• Applying Network Coding Algorithms to Broadcast Channels:

      Designing and implementing a new network coding algorithm for broadcast channel to reduce the required transmissions to deliver all packets to all devices. 

• Programming AVR Processors:

      Programming AVR processors to run some applications and control external devices. Moreover, we successfully installed a light version of Linux on ARM processor and played a song through a speaker connected to the processor.