WattsApp: Power-aware container scheduling

WattsApp: Power-aware container scheduling Mehta, Hemant Kumar; Harvey, Paul; Rana, Omar; Buyya, Rajkumar; Varghese, Blesson Containers are popular for deploying workloads. However, there are limited software-based methods (hardware-based methods are ex- pensive) for obtaining the power consumed by containers to facilitate power-aware container scheduling. This paper presents WattsApp, a tool underpinned by a six step software-based method for power- aware container scheduling to minimize power cap violations on a server. The proposed method relies on a neural network-based power estimation model and a power capped container scheduling technique. Experimental studies are pursued in a lab-based environment on 10 benchmarks on Intel and ARM processors. The results highlight that power estimation has negligible overheads - nearly 90% of all data samples can be estimated with less than a 10% error, and the Mean Absolute Percentage Error (MAPE) is less than 6%. The power-aware scheduling of WattsApp is more effective than Intel’s Running Power Average Limit (RAPL) based power capping as it does not degrade the performance of all running containers.
http://hdl.handle.net/10468/10681

DASH QoE performance evaluation framework with 5G datasets

DASH QoE performance evaluation framework with 5G datasets Ul Mustafa, Raza; Islam, Md. Tariqul; Rothenberg, Christian E.; Ferlin, Simone; Raca, Darijo; Quinlan, Jason J. Fifth Generation (5G) networks provide high throughput and low delay, contributing to enhanced Quality of Experience (QoE) expectations. The exponential growth of multimedia traffic pose dichotomic challenges to simultaneously satisfy network operators, service providers, and end-user expectations. Building QoE-aware networks that provide run-time mechanisms to satisfy end-users’ expectations while the end-to end network Quality of Service (QoS) varies is challenging and motivates many ongoing research efforts. The contribution of this work is twofold. Firstly, we present a reproducible data-driven framework with a series of pre-installed Dynamic Adaptive Streaming over HTTP (DASH) tools to analyse state of-art Adaptive Bitrate Streaming (ABS) algorithms by varying key QoS parameters in static and mobility scenarios. Secondly, we introduce an interactive Binder notebook providing a live analytical environment which processes the output dataset of the framework and compares the relationship of five QoE models, three QoS parameters (RTT, throughput, packets), and seven video KPIs.
http://hdl.handle.net/10468/10672

A supervised machine learning approach for DASH video QoE prediction in 5G networks

A supervised machine learning approach for DASH video QoE prediction in 5G networks Ul Mustafa, Raza; Ferlin, Simone; Rothenberg, Christian E.; Raca, Darijo; Quinlan, Jason J. Future fifth generation (5G) networks are envisioned to provide improved Quality-of-Experience (QoE) for applications by means of higher data rates, low and ultra-reliable latency and very high reliability. Proving increasing beneficial for mobile devices running multimedia applications. However, there exist two main co-related challenges in multimedia delivery in 5G. Namely, balancing operator provisioning and client expectations. To this end, we investigate how to build a QoE-aware network that guarantees at run-time that the end-to-end user experience meets the end users’ expectations at the same that the network’s Quality of Service (QoS) varies. The contribution of this paper is twofold: First, we consider a Dynamic Adaptive Streaming over HTTP (DASH) video application in a realistic emulation environment derived from real 5G traces in static and mobility scenarios to assess the QoE performance of three state-of-art Adaptive Bitrate Streaming (ABS) algorithm categories: Hybrid - Elastic and Arbiter+; buffer-based - BBA and Logistic; and rate-based - Exponential and Conventional. Second, we propose a Machine Learning (ML) classifier to predict user satisfaction which considers network metrics, such as RTT, throughput, and number of packets. Our proposed model does not rely on knowledge about the application or specific traffic information. We show that our ML classifiers achieves a QoE prediction accuracy of 87.63 % and 79 % for static and mobility scenarios, respectively.
http://hdl.handle.net/10468/10646

Scaling-invariant maximum margin preference learning

Scaling-invariant maximum margin preference learning Montazery, Mojtaba; Wilson, Nic One natural way to express preferences over items is to represent them in the form of pairwise comparisons, from which a model is learned in order to predict further preferences. In this setting, if an item a is preferred to the item b, then it is natural to consider that the preference still holds after multiplying both vectors by a positive scalar (e.g., ). Such invariance to scaling is satisfied in maximum margin learning approaches for pairs of test vectors, but not for the preference input pairs, i.e., scaling the inputs in a different way could result in a different preference relation being learned. In addition to the scaling of preference inputs, maximum margin methods are also sensitive to the way used for normalizing (scaling) the features, which is an essential pre-processing phase for these methods. In this paper, we define and analyse more cautious preference relations that are invariant to the scaling of features, or preference inputs, or both simultaneously; this leads to computational methods for testing dominance with respect to the induced relations, and for generating optimal solutions (i.e., best items) among a set of alternatives. In our experiments, we compare the relations and their associated optimality sets based on their decisiveness, computation time and cardinality of the optimal set.
http://hdl.handle.net/10468/10802