A Predictive and QoS Aware Kubernetes Container Placement Framework for Heterogeneous Mobile Edge Cloud Networks

Evolution and advancement of mobile computing, Virtual Reality (VR), IoT, Smart Vehicles and other related technologies and devices that need to be connected to the network all the time has increased over years. Many mobile devices have limited computation, storage and battery capacity to run applications. Some mobile applications require low latency and fast response time that can be achieved by offloading their compute workload to cloud networks. Since the datacentres hosting these cloud services are usually at distant locations far from the clients and the latency of such networks are high, these clouds cannot provide real-time services for some circumstances. Deploying clouds in the edge provides low latency but faces challenges of meeting user’s QoS requirements when they are mobile. In this thesis I proposed a new QoS-aware and predictive container placement framework for Kubernetes in edge cloud environments to serve mobile users and devices.