Abstract
This paper presents CoopREP, a system that provides support for fault replication of concurrent programs based on cooperative recording and partial log combination. CoopREP uses partial logging to reduce the amount of information that a given program instance is required to store to support deterministic replay. This allows reducing substantially the overhead imposed by the instrumentation of the code, but raises the problem of finding a combination of logs capable of replaying the fault. CoopREP tackles this issue by introducing several innovative statistical analysis techniques aimed at guiding the search of the partial logs to be combined and needed for the replay phase. CoopREP has been evaluated using both standard benchmarks for multithreaded applications and real-world applications. The results highlight that CoopREP can successfully replay concurrency bugs involving tens of thousands of memory accesses, while reducing recording overhead with respect to state-of-the-art noncooperative logging schemes by up to 13x (and by 2.4x on average).

Abstract
Computer-based modelling tools allow students to express their theories in models that can be simulated. In this way, students can use their theories operationally, confronting themselves with the consequences of their ideas. The ability of students to form and express a mental model will be expanded if they are given an opportunity to become aware of their own mental model by expressing this same model and comparing it to other models, like a consensus model. The building of numerical models of biophysical phenomena, such as the mechanics of breathing, or blood circulation, has the potential for student motivation as well as long-term learning. Our theory is that by re-building well known numerical models of physiological phenomena students will have the opportunity to change their perceptions about the relevance of the contents addressed, simultaneously improving their learning in the topics covered and increasing their motivation in the basic science disciplines in their curricula. For the implementation of computer numerical models historically it was necessary to use some programming language, such as MATLAB, BASIC, C++, JAVA. With the development of computer science, it is now possible to these students "construct" models of physical phenomena expressed through dedicated computer tools without necessarily having to do so in a programming language. As for example we have STELLA, MODELLUS, or STARLOGO. There is also the possibility of using a spreadsheet such as Microsoft EXCEL, Open Office CALC, Google SHEETS, or others, as tools that allow students to express physical models. The current spreadsheets, even those available for free, are very powerful, having many integrated tools, in terms of calculation, and we can count on several other features, such as graphs of various types, buttons and other tools that allow interaction with the model, and databases that can be integrated into the spreadsheet. There are several advantages of using spreadsheets in science education due to its general access, through smartphones, tablet's and computers, ease of implementation for the basic operations, and ease of the "debug" process, relative to other types of software. Also, it does not require prior knowledge of programming languages, or about complex mathematical software, which would an obstacle to the learning in itself. There is also the positive side effect of learning how to use a spreadsheet that is a plus in itself for the future professional's. This paper will have a review of the state of the art of using spreadsheets in Modelling Based Learning. Also, it will be presented a study with first year undergraduate students of a health care course, using Biophysical models historically very important in the physiology and medicine development.

Abstract
Network data volumes have seen a substantial increase in recent years, in part due to the massive use of mobile devices, the dissemination of streaming services and the rise of concepts such as IoT. This growing trend highlights the need to improve network monitoring systems to cope with challenges related with performance, flexibility and security. Software-Defined Networking (SDN) and traffic sampling techniques can be combined to provide a toolset that can be used for enhancing network management activities and performance evaluation. In this context, this paper presents a proposal for supporting time-based sampling techniques in SDN, providing network statistics at the controller level and allowing the self-configuration of traffic sampling in network devices. The proposed solution, designed to improve the efficiency and flexibility of network measurement systems, takes into account the underlying need to establish a balance between the reliability of the collected data and the computational effort involved in the sampling process. The proof-of-concept results emphasize the potential of applying and configuring different time-based sampling techniques through a SDN framework and a small set of standard OpenFlow messages. Comparative results on the accuracy and overhead of each technique when sampling real traffic traces are also provided.

Abstract
The multitude of Wireless Sensor Networks (WSNs) environments, being typically resource-constrained, clearly benefit from properties such as adaptiveness and energy-awareness, in particular, in presence of demanding data gathering applications. This paper proposes a self-adaptive, energy-aware sensing scheme for WSNs (e-LiteSense), which aims at self-adjusting the data gathering process to each specific WSN context, capturing accurately the behaviour of physical parameters of interest yet reducing the sensing overhead. The adaptive scheme relies on a set of low-complexity rules capable of auto-regulate the sensing frequency according to the parameters variability and energy levels. The proof-of-concept resorts to real-world datasets to provide evidence of e-LiteSense ability to optimise the data gathering process according to energy levels, improving the trade-off between accuracy and WSN lifetime.

Abstract

Abstract
Computer-aided engineering simulations, in particular, Computational Fluid Dynamics, have become a fundamental design and analysis tool in product development. Over time, a demand for larger problem sizes and higher accuracy has led to huge computational workloads requiring extended compute capabilities. Increasing computing capabilities requirements, however, drive a fast-growing power consumption. In order to deal with increasing power demand, hardware and software solutions' reevaluation in terms of power-efficiency becomes of paramount importance. Establishing a power budget and reducing the compute units operating frequency in order to comply with such budget is becoming common practice. However, in the presence of heterogeneous compute units and dynamic workloads, a static and uniform reduction across compute units leads to a potentially severe impact on performance. This paper proposes a run-time heterogeneity-aware power-adaptive schedule that provides power consumption optimization, targeting heterogeneous parallel distributed systems in the context of CFD simulations. The proposed approach is integrated into OpenFOAM computational library and explores power migration and reduction across nodes, considering runtime workload imbalances and node performances. Results reveal not only a substantial reduction in power usage but also significant performance gains relative to the uniform static approach. To the best of authors' knowledge, this is the first implementation and integration of power management solutions in OpenFOAM.