Multi-Hop Communications in the THz Band
This thesis focuses on the fundamental physical layer modeling of multi-hop communications in the THz band and corresponding system optimization. An intensive quantitative modeling and analysis is necessary to demonstrate the performance improvement of multi-hop links compared to the conventional single-hop case. Specifically, the following tasks related to system modeling will be tackled:
- Identify several practical use cases with detailed configurations to deploy multi-hop THz communications (e.g., Manhattan grid-type urban scenarios).
- Geometry-based statistical channel modeling on above identified real-world communication scenarios (use ray-tracing technique for numerical result validation).
Based on the developed system model, the tradeoff of energy consumption vs. number of hops will be optimized to combat the distance limitation at THz band and improve the spectral efficiency to be able to achieve a Terabits-per-second link throughput:
- Form the optimization problem and use e.g. linear programming approach for solution.
- Explore machine learning algorithms, such as reinforcement learning, to solve the same problem and compare different algorithms’ time complexity.
Guidelines for the project
- Conduct a literature survey on multi-hop communications in higher frequency bands
- Develop a system model as described above for multi-hop THz communications
- Optimize the transmission policy
- Conduct simulations for a performance evaluation of the proposed scheme
Prerequisites
- Comprehensive knowledge of digital communications
- Basic knowledge of machine learning
- Strong theoretical and mathematical background
- Experience in Matlab and Python
Contact