Context-sensitive Offloading for Genetic Algorithms in Heterogeneous, Distributed SystemsBachelor Thesis
16 January 2025, by Jonas Müller

Photo: mistral
With the steadily increasing digitalization, the volume of data to be processed as well as the spread of networked end devices with unused computing capacity is growing. This bachelor thesis investigates how such excess resources in a heterogeneous edge environment can be efficiently utilized to solve optimization problems. The focus of the analysis will be on the Traveling Salesman Problem (TSP), which will be examined using various algorithms such as the Brute-Force Algorithm, the Genetic Algorithm, and the Nearest-Neighbor Algorithm. Different termination criteria as well as parallelization strategies will be considered with the aim of optimizing calculations through distributed offloading. Finally, a client-side implementation is to be developed, with a focus on a combination of efficiency, parallelizability, and realistic evaluation.
Supervised by:
Prof. Dr. Janick Edinger, Anton Semjonov