Thesis topic: Amoeba-Inspired SAT Solver for IoT Edge Devices

This work aims at realizing a Satisfiability problem (SAT) solver, AmoebaSAT, whose algorithm was inspired from a single cell amoeba, on resource-constrained IoT edge devices. As the solver possesses a massive parallelism feature where processing elements perform computations in parallel to decide variable assignments, we conducted its FPGA-based hardware implementations using a high-level design approach. Based on the original AmoebaSAT solver, we extended the algorithm to help the solver escape from local minima more quickly and achieved significant speedup to outperform state-of-the-art hardware SAT solvers. Moreover, by utilizing the inter-variable dependency feature of different SAT instances, we achieved further iteration reduction without incurring effects on hardware usage. 

Thesis topic: High-level Synthesis of Amoeba-inspired SAT Solvers on FPGA

The Amoeba-inspired SAT (satisfiability) algorithm bases on a computing model inspired from the behavior of an amoeba (a single cell organism) in its adaption with the surrounding environment. This computing model, AmoebaSAT, can effectively solve SAT problems, which can be extended to understand various combinatorial optimization problems that appear in many practical applications, especially IoT applications. This work aims at realizing an efficient hardware AmoebaSAT solver on an FPGA, through exploring several hardware implementations by high level synthesis (HLS).

Thesis topic: Level of Traffic Service Classification System

Traffic Level of Service (LOS) information is crucial for traffic management systems, especially in urban areas. This work proposes a cost effective distributed traffic estimation solution using smart cameras, each of which is equipped with a low cost ARM microprocessor to estimate the LOS from the captured traffic images.