There is a growing need to manage water resources in a sustainable way, particularly in semi arid areas, with dramatic social and economic development as well as rapid population growth. Optimising water allocation in a river basin is an important aspect ensuring equitable and efficient water use. This research develops an optimisation approach (the Integrated Water Resource Optimisation model, IWRO) to optimise the conjunctive use of surface water and groundwater resources in a sustainable manner. The IWRO model is comprised of a surface water optimisation model (SWO) and the Tsinghua groundwater optimisation (TGO) model. These models employ Genetic Algorithms (GAs) to optimise water allocation. Application of a surface water optimisation (SWO) model incorporating a GA is demonstrated initially for a simple test case, through which the GA approach was validated against known solutions. Sensitivity analysis of different operators and parameters related to GAs was also carried out. The validated SWO model was then applied to a more complex system, the Shiyang River Basin in Gansu Province in China, to maximise equitable surface water supplies. On the groundwater side, the GA approach was applied with the existing Tsinghua groundwater model to optimise groundwater supplies with sustainability considerations. The results were compared with those from an existing model (the WEAP model), indicating that the IWRO model is capable of satisfying the objectives of equitable water allocation and groundwater sustainability set for it. In the context of Integrated Water Resources Management (IWRM), account must be taken of a wide range of social and environmental issues. Different scenarios were therefore designed for the Shiyang River Basin management. Various criteria in terms of economic, social, environment and water security were also indentified for further multi-criterion decision making analysis.

Utility Computing is one of the most discussed business models in the context of Cloud Computing. Service providers are more and more pushed into the role of utilities by their customer's expectations. Subsequently, the demand for predictable service availability and pay-per-use pricing models increases. Furthermore, for providers, a new opportunity to optimise resource usage offers arises, resulting from new virtualisation techniques. In this context, the control of service quality and profit depends on a deep understanding of the representation of the relationship between business and technique. This research analyses the relationship between the business model of Utility Computing and Service-oriented Computing architectures hosted in Cloud environments. The relations are clarified in detail for the entire service life cycle and throughout all architectural layers. Based on the elaborated relations, an approach to a delivery framework is evolved, in order to enable the optimisation of the relation attributes, while the service implementation passes through business planning, development, and operations. Related work from academic literature does not cover the collected requirements on service offers in this context. This finding is revealed by a critical review of approaches in the fields of Cloud Computing, Grid Computing, and Application Clusters. The related work is analysed regarding appropriate provision architectures and quality assurance approaches. The main concepts of the delivery framework are evaluated based on a simulation model. To demonstrate the ability of the framework to model complex pay-per-use service cascades in Cloud environments, several experiments have been conducted. First outcomes proof that the contributions of this research undoubtedly enable the optimisation of service quality and profit in Cloud-based Service-oriented Computing architectures.

A key challenge in the field of cloud computing is to automatically scale software containers in a way that accurately matches the demand for the services they run. To manage such components, container orchestrator tools such as Kubernetes are employed, and in the past few years, researchers have attempted to optimise its autoscaling mechanism with different approaches. Recent studies have showcased the potential of Actor-Critic Deep Reinforcement Learning (DRL) methods in container orchestration, demonstrating their effectiveness in various use cases. However, despite the availability of solutions that integrate multiple container performance metrics to evaluate autoscaling decisions, a critical gap exists in understanding how model-free DRL algorithms interact with a state space based on those metrics. Thus, the primary objective of this thesis is to investigate the impact of the state space definition on the performance of model-free DRL methods in the context of horizontal autoscaling within Kubernetes clusters. In particular, our findings reveal distinct behaviours associated with various sets of metrics. Notably, those sets that exclusively incorporate parameters present in the reward function demonstrate superior effectiveness. Furthermore, our results provide valuable insights when compared to related works, as our experiments demonstrate that a careful metric selection can lead to remarkable Service Level Agreement (SLA) compliance, with as low as 0.55% violations and even surpassing baseline performance in certain scenarios. / En viktig utmaning inom området molnberäkning är att automatiskt skala programvarubehållare på ett sätt som exakt matchar efterfrågan för de tjänster de driver. För att hantera sådana komponenter, container orkestratorverktyg som Kubernetes används, och i det förflutna några år har forskare försökt optimera dess autoskalning mekanism med olika tillvägagångssätt. Nyligen genomförda studier har visat potentialen hos Actor-Critic Deep Reinforcement Learning (DRL) metoder i containerorkestrering, som visar deras effektivitet i olika användningsfall. Men trots tillgången på lösningar som integrerar flera behållarprestandamått att utvärdera autoskalningsbeslut finns det ett kritiskt gap när det gäller att förstå hur modellfria DRLalgoritmer interagerar med ett tillståndsutrymme baserat på dessa mätvärden. Det primära syftet med denna avhandling är alltså att undersöka vilken inverkan statens rymddefinition har på prestandan av modellfria DRL-metoder i samband med horisontell autoskalning inom Kubernetes-kluster. I synnerhet visar våra resultat distinkta beteenden associerade med olika uppsättningar mätvärden. Särskilt de set som uteslutande innehåller parametrar som finns i belöningen funktion visar överlägsen effektivitet. Dessutom våra resultat ge värdefulla insikter jämfört med relaterade verk, som vår experiment visar att ett noggrant urval av mätvärden kan leda till anmärkningsvärt Service Level Agreement (SLA) efterlevnad, med så låg som 0, 55% överträdelser och till och med överträffande baslinjeprestanda i vissa scenarier.