Postdoctoral researcher, Trinity College Dublin
Service Aware Storage/Computing/Networking Resource Orchestration in Hybrid SDN Network Architecture
Today, we witness unprecedented changes in the landscape of wireless networks which challenge the performance and even the economic viability of Mobile Network Operators (MNOs). We see the explosive growth of mobile data traffic and, at the same time, the continuous deployment of novel communication and computing services with increasingly stringent performance requirements. These applications depart significantly from the voice-only and low data rate services of the past; they require huge bandwidth chunks and ultra low-delay data transfers, and involve computing and storage resources that are often owned by third-party Over-The-Top (OTT) service providers. These developments have spurred a flurry of industry and academic efforts, broadly known as the research for 5G networks and beyond (5G+), which aspire to design a new generation of wireless systems.
One very innovative feature of 5G+ network is the deployment of resources at the network edge. Prominent examples include the placement of small base stations in proximity with users, content caches at the network core or base stations, and deployment of computing servers at the last mile. Such solutions are oftentimes the only realistic way to support demanding requests, e.g., augmented reality apps or energy-prudent content transfers, and can substantially reduce network costs. The significance of these edge-networking (EN) architectures is well-recognized, as is manifested by the establishment of industry-academia consortiums such as the OpenFog and MEC, and the launch of dedicated conferences. However, we currently lack the mechanisms for realising such systems, since there are certain bottleneck issues that have to be addressed: 1) Resources at the edge are not a panacea. 2) Lack of service-aware performance metrics. 3) Development of novel business models in edge networking. Our goal in this research is to address the above open issues by delivering a multi-faceted framework that enables service-aware resource orchestration in hybrid cloud/edge network architecture (SERHENA).
To this end, the main objectives of this research are 1) dimensioning and resource management policies, 2) mechanisms for workload and content sharing at the edge, 3) economics of service-aware edge-networking.
Abstract: We consider the problem of power demand forecasting in residential micro-grids. Several approaches using ARMA models, support vector machines, and recurrent neural networks that perform one-step ahead predictions have been proposed in the literature. Here, we extend them to perform multi-step ahead forecasting and we compare their performance. Toward this end, we implement a parallel and efficient training framework, using power demand traces from real deployments to gauge the accuracy of the considered techniques. Our results indicate that machine learning schemes achieve smaller prediction errors in the mean and the variance with respect to ARMA, but there is no clear algorithm of choice among them. Pros and cons of these approaches are discussed and the solution of choice is found to depend on the specific use case requirements. A hybrid approach, that is driven by the prediction interval, the target error, and its uncertainty, is then recommended.
Pub.: 29 Jun '17, Pinned: 26 Aug '17
Abstract: The unstable network connectivity is the bottleneck of providing Gaming as a service (GaaS) for mobile devices. Therefore, the most critical technical challenge is to compress and transmit the real-time gaming video, so that during the gaming session, the expected server transmission rate over the bandwidth-limited mobile network can be minimized, while satisfying the quality of experience for the players. Inspired by the idea of peer-to-peer sharing between multiple players, we propose a cloudlet-assisted multiplayer cloud gaming system, in which the mobile devices are connected to the cloud server for real-time interactive game videos, while sharing the received video frames with their peers via an ad hoc cloudlet. Experimental results show that expected server transmission rate can be significantly reduced compared to the conventional video encoding schemes for cloud games.
Pub.: 09 Nov '13, Pinned: 26 Aug '17