Document VersionPublisher's PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:• A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publicationCitation for published version (APA): Gheorghita, S. V., Palkovic, M., Hamers, J., Vandecappelle, A., Mamagkakis, S., Basten, T., ... Bosschere, . System scenario based design of dynamic embedded systems. (ES reports; Vol. 2007-06). Eindhoven: Technische Universiteit Eindhoven. General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.• You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ? Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. In the past decade, real-time embedded systems have become much more complex due to the introduction of a lot of new functionality in one application, and due to running multiple applications concurrently. This increases the dynamic nature of today's applications and systems, and tightens the requirements for their constraints in terms of deadlines and energy consumption. State-of-theart design methodologies try to cope with these novel issues by identifying several most used cases and dealing with them separately, reducing the newly introduced complexity. This paper presents a generic and systematic design-time/run-time methodology for handling the dynamic nature of modern embedded systems, which can be utilized by existing design methodologies to increase their efficiency. It is based on the concept of system scenarios, which group system behaviors that are similar from a multi-dimensional cost perspective, such as resource requirements, delay, and energy consumption, in such a way that the system can be configured to exploit this cost similarity. At design-time, these scenarios are individually optimized. Mechanisms for predicting the current scenario at run-time and for switching between scenarios are ...

Abstract. Most programs are repetitive, meaning that some parts of a program are executed more than once. As a result, a number of phases can be extracted in which each phase exhibits similar behavior. These phases can then be exploited for various purposes such as hardware adaptation for energy efficiency. Temporal phase classification schemes divide the execution of a program into consecutive (fixed-length) intervals. Intervals showing similar behavior are grouped into a phase. When a temporal scheme is used in an on-line system, phase predictors are necessary to predict when the next phase transition will occur and what the next phase will be. In this paper, we analyze and compare a number of existing state-of-the-art phase predictors using the SPEC CPU2000 benchmarks. The design space we explore is huge. We conclude that the 2-level burst predictor with confidence and conditional update is today's most accurate phase predictor within reasonable hardware budgets.

We consider the problem of developing suitable learning representations (embeddings) for library packages that capture semantic similarity among libraries. Such representations are known to improve the performance of downstream learning tasks (e.g. classification) or applications such as contextual search and analogical reasoning.We apply word embedding techniques from natural language processing (NLP) to train embeddings for library packages ("library vectors"). Library vectors represent libraries by similar context of use as determined by import statements present in source code. Experimental results obtained from training such embeddings on three large open source software corpora reveals that library vectors capture semantically meaningful relationships among software libraries, such as the relationship between frameworks and their plug-ins and libraries commonly used together within ecosystems such as big data infrastructure projects (in Java), front-end and back-end web development frameworks (in JavaScript) and data science toolkits (in Python).

Cyber-Physical Systems (CPSs) are complex systems comprising computation, physical, and networking assets. Used in various domains such as manufacturing, agriculture, vehicles, etc., they blend the control of the virtual and physical worlds. Smart homes are a peculiar type of CPS where the local networking fundamentals have seen little evolution in the past decades, while the context in which home networks operate has drastically evolved. With the advent of the Internet of Things (IoT), the number and diversity of devices connected to our home networks are exploding. Some of those devices are poorly secured and put users’ data privacy and security at risk. At the same time, administrating a home network has remained a tedious chore, requiring skills from un-savvy users. We present Future Spaces, an end-to-end hardware-software prototype providing fine-grained control over IoT connectivity to enable easy and secure management of smart homes. Relying on Software-Defined Networking-enabled home gateways and the virtualization of network functions in the cloud, we achieve advanced networking security and automation through the definition of isolated, usage-oriented slices. This disrupts how users discover, control and share their connected assets across multiple domains, smoothly adapting to various usage contexts.

-This paper introduces a new paradigm for service oriented networking being developed in the FUSION project 1 . Despite recent proposals in the area of information centric networking, a similar treatment of services -where networked software functions, rather than content, are dynamically deployed, replicated and invoked -has received little attention by the network research community to date. Our approach provides the mechanisms required to deploy a replicated service instance in the network and to route client requests to the closest instance in an efficient manner. We address the main issues that such a paradigm raises including load balancing, resource registration, domain monitoring and inter-domain orchestration. We also present preliminary evaluation results of current work.

The visualization tool rdvis is presented which aims at helping the programmer to find program transformations to improve temporal data locality. We present a number of locality metrics that capture the necessary information. Based on a cluster analysis of basic block vectors, the tool gives strong hints on which program transformations are needed. The visualizer allowed us to find the necessary transformations for three SPEC2000 programs in just a few minutes. After performing these transformations, the programs run 3 times faster on average on a number of different platforms.

Abstract-Optimal placement and selection of service instances in a distributed heterogeneous cloud is a complex trade-off between application requirements and resource capabilities that requires detailed information on the service, infrastructure constraints and the underlying IP network. In this article we first posit that from an analysis of a snapshot of today's centralized and regional data centre infrastructure, there is a sufficient number of candidate sites for deploying many services while meeting latency and bandwidth constraints. We then provide quantitative arguments why both network and hardware performance needs to be taken into account when selecting candidate sites to deploy a given service. Lastly, we propose a novel architectural solution for service-centric networking. The resulting system exploits the availability of fine-grained execution nodes across the Internet and uses knowledge of available computational and network resources for deploying, replicating and selecting instances to optimize Quality of Experience for a wide range of services. I. INTERACTIVE DEMANDING SERVICES IN THE CLOUDThere is vast diversity in cloud-hosted services today, ranging from mobile back-ends, over virtualized set-top boxes and gaming consoles to real-time services providing decision and control support for self-driving cars. These recent cloud services require a crisp experience and/or realtime processing of high data rate streams. High network delays and low throughput to a relatively small number of centralised remote data centres (DCs) may have a serious impact on the quality of experience (QoE). For instance, 30% of the US population has a too high latency to one of Amazon's EC2 DCs for cloud-based gaming [1]. Deploying such applications in distributed execution platforms closer to the users reduces network delays and is also the preferred approach for many data intensive applications. Shifting all the data to a centralised service could overwhelm the network and it is better to bring the computation logic closer to data sources and users at the network edge. As of today, Internet Service Providers (ISPs) already deploy Content Delivery Network (CDN) proxy servers in their network to save on transit costs and improve the quality of service for their customers [2].Service developers are thus confronted with the twofold challenge of service instance placement and selection. The central problem in service placement is to determine the cost-optimal set of geo-distributed datacenters where to deploy an instance, and to configure the appropriate scaling policies in each of these datacenters to adequately cope with the expected demand. These distributed nodes have heterogeneous hardware, as they are owned by different entities or deployed at different moments in time. Service instance selection refers to the anycast-style resolution of a service identifier to the network endpoint of the best replica, taking into account service availability, network metrics and the location of the requesting user. Service placemen...

Abstract-Today's centralized cloud-computing infrastructures have not been designed with geo-localized, personalized, bandwidth/processing-intensive, real-time applications in mind. High network delay and low throughput can have a significant impact on the user experience. Instead, such services could be deployed in distributed service nodes at the edge of the network, closer to the user. In this paper we focus on composite services of which the components are running in different service nodes. We present a two-layer framework that provides service orchestration and instance selection. We present the orchestration mechanisms to enable the flexible re-use of components across different composite services. For the resolution layer of our framework, we present two modes of operation that combine network and service availability information for efficient per-request instance selection among a multitude of service replicas.