Abstract:
Resource management is a generic term for the process of determining the most beneficial way to employ some limited set of resources. Typically, not all demands for a set of resources can be fulfilled and determining their most efficient usage is challenging. To make matters worse, many problems are not static. Instead, the state of a system changes dynamically and depends on the usage of its resources in time. Planning the usage of resources in time is denoted as scheduling in the literature. Computing schedules is hard and requires the accurate modelling of the underlying system.
The objective of this work is the development and evaluation of scheduling algorithms for novel applications in power systems, real-time communication, and multicast protocols. We contribute various research works about scheduling in power systems. We propose a scheduling algorithm to save energy costs in production processes, evaluate forecasting methods for domestic demands, and present an optimization algorithm to maximize self-consumption of roof-top PV energy with a heat pump. In the domain of real-time communication, we review the state of the art of scheduling algorithms for the so-called Time-Aware Shaper (TAS). We report results of a quantitative study comparing various scheduling algorithms for TAS. Additionally, we propose a scheduling algorithm that computes schedules robust against some sources of non-determinism. In the field of multicast protocols, we first contribute a partitioning algorithm needed to scale Bit Indexed Explicit Replication (BIER) to large networks. Then, we present Segment-Encoded Explicit Trees (SEET), a novel multicast protocol that allows tree engineering.
The research presented in this thesis has been funded by different research projects by the Deutsche Forschungsgemeinschaft (DFG) under grant ME2727/1-2, the German Federal Ministry of Education and Research (BMBF) under support code 16KIS1161 (Collaborative Project KITOS), and the German Federal Ministry for Economic Affairs and Energy 16KN039521 (ZIM).