Search

Radio links in wireless body area networks (WBANs) are highly subject to short and long-term attenuation due to the unstable network topology and frequent body blockage. This instability makes it challenging to achieve reliable and energy-efficient communication, but on the other hand, provides a great potential for the sending nodes to dynamically schedule the transmissions at the time with the best expected channel quality. Motivated by this, we propose improved Gilbert-Elliott Markov chain model (IGE), a memory-efficient Markov chain model to monitor channel fluctuations and provide a long-term channel prediction. We then design adaptive transmission power selection (ATPS), a deadline-constrained channel scheduling scheme that enables a sending node to buffer the packets when the channel is bad and schedule them to be transmitted when the channel is expected to be good within a deadline. ATPS can self-learn the pattern of channel changes without imposing a significant computation or memory overhead on the sending node. We evaluate the performance of ATPS through experiments using TelosB motes under different scenarios with different body postures and packet rates. We further compare ATPS with several state-of-the-art schemes, including the optimal scheduling policy, in which the optimal transmission time for each packet is calculated based on the collected received signal strength indicator (RSSI) samples in an off-line manner. The experimental results reveal that ATPS performs almost as efficiently as the optimal scheme in high-date-rate scenarios and has a similar trend on power level usage.

The advent of edge devices dedicated to machine learning tasks enabled the execution of AI-based applications that efficiently process and classify the data acquired by the resource-constrained devices populating the Internet of Things. The proliferation of such applications (e.g., critical monitoring in smart cities) demands new strategies to make these systems also sustainable from an energetic point of view. In this paper, we present an energy-aware approach for the design and deployment of self-adaptive AI-based applications that can balance application objectives (e.g., accuracy in object detection and frames processing rate) with energy consumption. We address the problem of determining the set of configurations that can be used to self-adapt the system with a meta-heuristic search procedure that only needs a small number of empirical samples. The final set of configurations are selected using weighted gray relational analysis, and mapped to the operation modes of the self-adaptive application. We validate our approach on an AI-based application for pedestrian detection. Results show that our self-adaptive application can outperform non-adaptive baseline configurations by saving up to 81% of energy while loosing only between 2% and 6 % in accuracy.

New HPC machines are getting close to the exascale. Power consumption for those machines has been increasing, and researchers are studying ways to reduce it. A second trend is HPC machines' growing complexity, with increasing heterogeneous hardware components and different clusters architectures cooperating in the same machine. We refer to these environments with the term heterogeneous multi-cluster environments. With the aim of optimizing performance and energy consumption in these environments, this paper proposes an Energy-Aware-Multi-Cluster (EAMC) job scheduling policy. EAMC-policy is able to optimize the scheduling and placement of jobs by predicting performance and energy consumption of arriving jobs for different hardware architectures and processor frequencies, reducing workload's energy consumption, makespan, and response time. The policy assigns a different priority to each job-resource combination so that the most efficient ones are favored, while less efficient ones are still considered on a variable degree, reducing response time and increasing cluster utilization. We implemented EAMC-policy in Slurm, and we evaluated a scenario in which two CPU clusters collaborate in the same machine. Simulations of workloads running applications modeled from real-world show a reduction of response time and makespan by up to 25% and 6% while saving up to 20% of total energy consumed when compared to policies minimizing runtime, and by 49%, 26%, and 6% compared to policies minimizing energy., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project, by the Generalitat de Catalunya (2017-SGR-1414), from the European Union’s Horizon 2020 under grant agreement No. 754304 (DEEP-EST Project), and the BSC-Lenovo collaboration., Peer Reviewed, Postprint (author's final draft)

Hadoop, a pioneering open source framework, has revolutionized the big data world because of its ability to process vast amounts of unstructured and semi-structured data. This ability makes Hadoop the ‘go-to’ technology for many industries that generate big data, thus it also aids in being cost effective, unlike other legacy systems. Hadoop MapReduce is used in large scale data parallel applications to process massive amounts of data across a cluster and is used for scheduling, processing, and executing jobs. Basically, MapReduce is the right hand of Hadoop, as its library is needed to process these large data sets. In this research thesis, this study proposes a smart framework model that profiles MapReduce tasks with the use of Machine Learning (ML) algorithms to effectively place the data in Hadoop clusters; activate only sufficient number of nodes to accomplish the data processing within the planned deadline time for the task. The model will ensure achieving energy efficiency by utilizing the minimum number of necessary nodes, with maximum utilization and least energy consumption to reduce the overall cost of operations in data centers that deploy the Hadoop clusters.

Hadoop, a pioneering open source framework, has revolutionized the big data world because of its ability to process vast amounts of unstructured and semi-structured data. This ability makes Hadoop the ‘go-to’ technology for many industries that generate big data, thus it also aids in being cost effective, unlike other legacy systems. Hadoop MapReduce is used in large scale data parallel applications to process massive amounts of data across a cluster and is used for scheduling, processing, and executing jobs. Basically, MapReduce is the right hand of Hadoop, as its library is needed to process these large data sets. In this research thesis, this study proposes a smart framework model that profiles MapReduce tasks with the use of Machine Learning (ML) algorithms to effectively place the data in Hadoop clusters; activate only sufficient number of nodes to accomplish the data processing within the planned deadline time for the task. The model will ensure achieving energy efficiency by utilizing the minimum number of necessary nodes, with maximum utilization and least energy consumption to reduce the overall cost of operations in data centers that deploy the Hadoop clusters.

Collaboration between multiple Unmanned Aerial Vehicles (UAVs) to set up a Flying Ad Hoc Network (FANET) is a growing trend since future applications claim for more autonomous and rapid deployable systems. The user experience on watching videos transmitted over FANETs should always be satisfactory even under influence of topology changes caused by the energy consumption of UAVs. In addition, the FANET must keep the UAVs cooperating as much as possible during a mission. However, one of the main challenges in FANET is how to mitigate the impact of limited energy resources of UAVs on the FANET operation in order to monitor the environment for a long period of time. In this sense, UAV replacement is required in order to avoid the premature death of nodes, network disconnections, route failures, void areas, and low-quality video transmissions. In addition, decision-making must take into account energy consumption associated with UAV movements, since they are generally quite energy-intensive. This article proposes a cooperative UAV scheme for enhancing video transmission and global energy efficiency called VOEI. The main goal of VOEI is to maintain the video with QoE support while supporting the nodes with a good connectivity quality level and flying for a long period of time. Based on an Software Defined Network (SDN) paradigm, the VOEI assumes the existence of a centrailized controller node to compute reliable and energy-efficiency routes, as well as detects the appropriate moment for UAV replacement by considering global FANET context information to provide energy-efficiency operations. Based on simulation results, we conclude that VOEI can effectively mitigate the energy challenges of FANET, since it provides energy-efficiency operations, avoiding network death, route failure, and void area, as well as network partitioning compared to state-of-the-art algorithm. In addition, VOEI delivers videos with suitable Quality of Experience (QoE) to end-users at any time, which

Collaboration between multiple Unmanned Aerial Vehicles (UAVs) to set up a Flying Ad Hoc Network (FANET) is a growing trend since future applications claim for more autonomous and rapid deployable systems. The user experience on watching videos transmitted over FANETs should always be satisfactory even under influence of topology changes caused by the energy consumption of UAVs. In addition, the FANET must keep the UAVs cooperating as much as possible during a mission. However, one of the main challenges in FANET is how to mitigate the impact of limited energy resources of UAVs on the FANET operation in order to monitor the environment for a long period of time. In this sense, UAV replacement is required in order to avoid the premature death of nodes, network disconnections, route failures, void areas, and low-quality video transmissions. In addition, decision-making must take into account energy consumption associated with UAV movements, since they are generally quite energy-intensive. This article proposes a cooperative UAV scheme for enhancing video transmission and global energy efficiency called VOEI. The main goal of VOEI is to maintain the video with QoE support while supporting the nodes with a good connectivity quality level and flying for a long period of time. Based on an Software Defined Network (SDN) paradigm, the VOEI assumes the existence of a centrailized controller node to compute reliable and energy-efficiency routes, as well as detects the appropriate moment for UAV replacement by considering global FANET context information to provide energy-efficiency operations. Based on simulation results, we conclude that VOEI can effectively mitigate the energy challenges of FANET, since it provides energy-efficiency operations, avoiding network death, route failure, and void area, as well as network partitioning compared to state-of-the-art algorithm. In addition, VOEI delivers videos with suitable Quality of Experience (QoE) to end-users at any time, which

Backbone telecommunication network infrastructures are deployed with redundant resources taking into account the backup capacity for protection in order to be resilient against link failures, and serving extremely large amount of data transmission resulting in increasing power consumption. In this study, the interplay between green and resilient network design, and flow routing mechanisms is analyzed. We propose Mixed Integer Linear Programming (MILP) models to obtain optimum solutions under various objectives: Minimizing consumption of (I) Capacity, (II) Capacity+ Power, and (III) Power. Two different shared backup protection (SBP) schemes (1) SBP-ind (failure independent) and (2) SBP-dep (failure dependent) are compared with dedicated path protection (DPP). It is assumed that network links utilized by only backup paths can be put into sleep mode. It is observed that when power consumption is minimized, the backup sharing decreases in SBP and, in the extreme case, it behaves similar to DPP. The models are generalized and valid for both IP traffic flow routing and lightpath routing. It is shown that for a sample network topology, to save e.g., 32.33% power, capacity consumption increases significantly, e.g., in SBP-ind up to 127.53%. In order to achieve a compromise between power and capacity consumption, we propose a multi-objective approach. All the MILP models are run and results are presented for a small scale European network topology as well as a larger scale sample US network topology. For larger problem instances ILP solutions are not scalable. Therefore, a novel energy efficient and survivable routing and network design algorithm, called energy-aware shared path protection (EASPP), addressing the trade-off caused by conflicting objectives of green and resilient network planning is proposed. Moreover this study presents a complete picture of various survivability mechanisms when power consumption is minimized together with the capacity consumption., QC 20170202

Today video streaming has become the biggest contributor to Internet data traffic. With the ever increasingpopularity of smart mobile devices for creating and consuming video content, in 5 years, mobile video contentis predicted to account for over 78% of total mobile data traffic. This popularity of mobile devices for mediaconsumption has resulted in a massive infrastructure deployment to support the growth. An adverse side effectof this growth is that IT industry is now a major contributor to total energy consumption in the world. Onthe other hand, users are struggling to manage the battery life of mobile devices as the battery technologyon these devices has not kept pace despite the advancements in processing power, connectivity and sensingcapabilities.This thesis introduces a novel energy efficient architecture for video streaming applications which factorsinto consideration the energy consumption of client devices as well as backend servers without negativelyimpacting the quality of experience of the end users. To this end, the conventional Quality of Experience(QoE) model in video streaming is augmented by incorporating energy status into the QoE model that hasled the development of eDASH, an energy-aware dynamic adaptive streaming over HTTP, chunk selectionalgorithm for video streaming players. Furthermore, feasibility of utilising low-power servers to host anddeliver video content is also investigated. Finally, this thesis proposes an energy efficient server managementframework and content pre-caching algorithm for efficient utilisation of low-power servers in conjunction witheDASH clients.

According to industry estimates, monthly global mobile data traffic will surpass 30.6 exabytes by 2020 and global mobile data traffic will increase nearly eightfold between 2015 and 2020. Most of the mobile data traffic is generated by smartphones, and the total number of smartphones is expected to continue growing by 2020, which results in rapid traffic growth. In addition, the upcoming 5G networks and Internet of Things based communication are estimated to involve a large amount of network traffic. The increase in mobile data traffic and in the number of connected devices poses a challenge to network operators, service providers, and data center operators. If the transmission capacity of the network and the amount of data traffic are not in line with each other, congestion may occur and ultimately the quality of experience degrades. Mobile networks are also becoming more reliant on data centers that provide efficient computing power. However, the energy consumption of data centers has grown in recent years, which is a problem for data center operators. A traditional strategy to overcome these problems is to scale up the resources or by providing more efficient hardware. Resource over-provisioning increases operating and capital expenditures without a guarantee of increased average revenue per user. In addition, the growing complexity and dynamics of communication systems is a challenge for efficient resource management. Intelligent and resilient methods that can efficiently use existing resources by making autonomous decisions without intervention from human administrators are thus needed. The goal of this research is to implement, develop, model, and test algorithms that can enable efficient and energy-aware network resource management in autonomous communications systems. First, an energy-aware algorithm is introduced for high-performance computing data centers to reduce the energy consumption within a single data center and across a federation of data, Tiivistelmä Langattoman tietoliikenteen nopean kasvun ennustetaan jatkuvan edelleen lähivuosinakin ja alan teollisuuden arvioiden mukaan matkapuhelinliikenteen määrä ylittäisi globaalisti 30,6 eksatavua vuoteen 2020 mennessä. Tämä tarkoittaisi liikennemäärän kahdeksankertaistumista ajanjaksolla 2015–2020. Älypuhelimet tuottavat suurimman osan matkapuhelinliikenteestä, ja älypuhelimien lukumäärän arvioidaan jatkavan kasvuaan vuoteen 2020 saakka, mikä johtaa nopeaan liikenteen kasvuun. Tämän lisäksi arvioidaan, että 5G verkot ja esineiden Internet tuottavat suuren määrän verkkoliikennettä. Matkapuhelinliikenteen ja laitteiden määrän kasvu tuo haasteita verkko-operaattoreille, palvelun tarjoajille, ja datakeskusoperaattoreille. Mikäli verkossa ei ole tarpeeksi siirtokapasiteettia dataliikenteen määrää varten, verkko ruuhkautuu ja lopulta palvelukokemus kärsii. Matkapuhelinverkot tulevat myös tulevaisuudessa tarvitsemaan datakeskusten laskentakapasiteettia. Datakeskusten energiankulutus on kuitenkin kasvanut viime vuosina, mikä on ongelma datakeskusoperaattoreille. Perinteinen strategia ongelmien ratkaisemiseksi on lisätä resurssien määrää tai tarjota tehokkaampaa laitteistoa. Resurssien liiallinen lisääminen kasvattaa kuitenkin sekä käyttö- että pääomakustannuksia ilman takuuta siitä, että keskimääräinen myyntitulo per käyttäjä kasvaisi. Tämän lisäksi tietoliikennejärjestelmät ovat monimutkaisia ja dynaamisia järjestelmiä, minkä vuoksi tehokas resurssienhallinta on haastavaa. Tämän vuoksi tarvitaan älykkäitä ja kestäviä metodeja, jotka pystyvät käyttämään olemassa olevia resursseja tehokkaasti tekemällä autonomisia päätöksiä ilman ylläpitäjän väliintuloa. Tämän tutkimuksen tavoitteena on toteuttaa, kehittää, mallintaa, ja testata algoritmeja, jotka mahdollistavat tehokkaan ja energiatietoisen verkkoresurssien hallinnan autonomisissa tietoliikennejärjestelmissä. Tutkimus esittää aluksi supertietokonedatakeskuksiin energiatietoisen algoritmin, jonka avulla voidaan vähent

The final publication is available at Springer via http://dx.doi.org/10.1007/s11227-014-1332-5, As the number of cores increases in current and future chip-multiprocessor (CMP) generations, coherence protocols must rely on novel hardware structures to scale in terms of performance, power, and area. Systems that use directory information for coherence purposes are currently the most scalable alternative. This paper studies the important differences between the directory behavior of private and shared blocks, which claim for a separate management of both types of blocks at the directory. We propose the PS directory, a two-level directory cache that keeps the reduced number of frequently accessed shared entries in a small and fast first-level cache, namely Shared cache, and uses a larger and slower second-level Private cache to track the large amount of private blocks. Entries in the Private cache do not implement the sharer vector, which allows important silicon area savings. Speed and area reasons suggest the use of eDRAM technology, much denser but slower than SRAM technology, for the Private cache, which in turn brings energy savings. Experimental results for a 16-core CMP show that, compared to a conventional directory, the PS directory improves performance by 14 % while reducing silicon area and energy consumption by 34 and 27 %, respectively. Also, compared to the state-of-the-art Multi-Grain Directory, the PS directory apart from increasing performance, it reduces power by 18.7 %, and provides more scalability in terms of area.

The final publication is available at Springer via http://dx.doi.org/10.1007/s11227-014-1332-5, As the number of cores increases in current and future chip-multiprocessor (CMP) generations, coherence protocols must rely on novel hardware structures to scale in terms of performance, power, and area. Systems that use directory information for coherence purposes are currently the most scalable alternative. This paper studies the important differences between the directory behavior of private and shared blocks, which claim for a separate management of both types of blocks at the directory. We propose the PS directory, a two-level directory cache that keeps the reduced number of frequently accessed shared entries in a small and fast first-level cache, namely Shared cache, and uses a larger and slower second-level Private cache to track the large amount of private blocks. Entries in the Private cache do not implement the sharer vector, which allows important silicon area savings. Speed and area reasons suggest the use of eDRAM technology, much denser but slower than SRAM technology, for the Private cache, which in turn brings energy savings. Experimental results for a 16-core CMP show that, compared to a conventional directory, the PS directory improves performance by 14 % while reducing silicon area and energy consumption by 34 and 27 %, respectively. Also, compared to the state-of-the-art Multi-Grain Directory, the PS directory apart from increasing performance, it reduces power by 18.7 %, and provides more scalability in terms of area.

We address a multicriteria nonpreemptive energy-aware scheduling problem for computationalGrid systems. This work introduces a new formulation of the scheduling problem for multicore heterogeneous computational Grid systems in which the minimization of the energy consumption, along with the makespan metric, is considered. We adopt a two-level model, in which a meta-broker agent (level 1) receives all user tasks and schedules them on the available resources, belonging to different local providers (level 2). The computing capacity and energy consumption of resources are taken from real multi-core processors from the main current vendors. Twenty novel list scheduling methods for the problem are proposed, and a comparative analysis of all of them over a large set of problem instances is presented. Additionally, a scalability study is performed in order to analyze the contribution of the best new bi-objective list scheduling heuristics when the problem dimension grows. We conclude after the experimental analysis that accurate trade-off schedules are computed by using the new proposed methods.

Recently, a broad range of ENS applications have appeared for large-scale systems, introducing new requirements leading to new embedded architectures, associated algorithms, and supporting software systems. These new requirements include the need for diverse and complex sensor systems that present demands for energy and computational resources as well as for broadband communication. To satisfy application demands while maintaining critical support for low energy operation, a new multiprocessor node hardware and software architecture, Low Power Energy Aware Processing (LEAP), has been developed. In this thesis we described the LEAP design approach, in which the system is able to adaptively select the most energy efficient hardware components matching an application's needs. The LEAP approach supports highly dynamic requirements in sensing fidelity, computational load, storage media, and network bandwidth. It focuses on episodic operation of each component and considers the energy dissipation for each platform task by integrating fine-grained energy dissipation monitoring and sophisticated power control scheduling for all subsystems, including sensors. In addition to LEAP's unique hardware capabilities, its software architecture has been designed to provide an easy to use power management interface, a robust, fault tolerant operating environment, and to enable remote upgrade of individual software components. Current research topics such as mobile computing and embedded networked sensing (ENS) have been addressing energy efficiency as a cornerstone necessity, due to their requirement for portability and long battery life times. This thesis discusses one such related project that, while currently directed toward ENS computing applications, is generally applicable to a wide ranging set of applications including both mobile and enterprise computing. While relevant to many applications, it is focuses on ENS environments necessitating high performance computing, networking

Recently, a broad range of ENS applications have appeared for large-scale systems, introducing new requirements leading to new embedded architectures, associated algorithms, and supporting software systems. These new requirements include the need for diverse and complex sensor systems that present demands for energy and computational resources as well as for broadband communication. To satisfy application demands while maintaining critical support for low energy operation, a new multiprocessor node hardware and software architecture, Low Power Energy Aware Processing (LEAP), has been developed. In this thesis we described the LEAP design approach, in which the system is able to adaptively select the most energy efficient hardware components matching an application's needs. The LEAP approach supports highly dynamic requirements in sensing fidelity, computational load, storage media, and network bandwidth. It focuses on episodic operation of each component and considers the energy dissipation for each platform task by integrating fine-grained energy dissipation monitoring and sophisticated power control scheduling for all subsystems, including sensors. In addition to LEAP's unique hardware capabilities, its software architecture has been designed to provide an easy to use power management interface, a robust, fault tolerant operating environment, and to enable remote upgrade of individual software components. Current research topics such as mobile computing and embedded networked sensing (ENS) have been addressing energy efficiency as a cornerstone necessity, due to their requirement for portability and long battery life times. This thesis discusses one such related project that, while currently directed toward ENS computing applications, is generally applicable to a wide ranging set of applications including both mobile and enterprise computing. While relevant to many applications, it is focuses on ENS environments necessitating high performance computing, networking

Energy consumption in optical backbone networks is increasing due to two main reasons: (i) the exponential growth of bandwidth demands, and (ii) the increase in availability requirements in order to guarantee protection of the ultra high capacity optical channels provided by wavelength division multiplexing (WDM) networks. Although state of the art reliability mechanisms are very efficient in guaranteeing high availability, they do not consider the impact of the protection resources on the network's energy consumption. Dedicated (1:1) path protection (DPP) is a well-known mechanism that provides one extra link-disjoint path for the protection of a connection request. This secondary path is reserved and maintained in an active mode even though it is not utilized most of the time. This means that in-line optical amplifiers and switching nodes/ports are always consuming power even when they are not used to reroute any primary traffic. Moreover secondary paths are on average longer than their respective primary paths. These observations motivated us to investigate the energy savings, when all unused protection resources can be switched into a low-power, stand-by state (or sleep mode) during normal network operation and can be activated upon a failure. It is shown that significant reduction of power consumption (up to 25%) can be achieved by putting protection resources into sleep mode. Moreover, in order to enhance this energy saving figure, this paper proposes and evaluates different energy-efficient algorithms, specifically tailored around the sleep mode option, to dynamically provision 1:1 dedicated path protected connection. The trade-off between energy saving and blocking probability is discussed and an efficient mechanism to overcome this drawback is devised. Our results reveal that a 34% reduction of energy consumption can be obtained with a negligible impact on the network's blocking performance.

QC 20130429

Scalability is a key design challenge that routing protocols for ad hoc networks must properly address to maintain the network performance when the number of nodes increases. We focus on this issue by reducing the amount of control information messages that a link state proactive routing algorithm introduces to the network. Our proposal is based on the observation that a high percentage of those messages is always the same. Therefore, we introduce a new mechanism that can predict the control messages that nodes need for building an accurate map of the network topology so they can avoid resending the same messages. This prediction mechanism, applied to OLSR protocol, could be used to reduce the number of messages transmitted through the network and to save computational processing and energy consumption. Our proposal is independent of the OLSR configuration parameters and it can dynamically self-adapt to network changes., Postprint (published version)

This paper describes a new embedded networked sensor platform architecture that combines hardware and software tools providing detailed, fine-grained real-time energy usage information. We introduce the LEAP2 platform, a qualitative step forward over the previously developed LEAP and other similar platforms. LEAP2 is based on a new low power ASIC system and generally applicable supporting architecture that provides unprecedented capabilities for directly observing energy usage of multiple subsystems in real-time. Real-time observation with microsecond-scale time resolution now enables direct accounting of energy dissipation for each computing task as well as for each hardware subsystem. This new hardware architecture is exploited with our new software tools, etop and endoscope. A series of experimental investigations provide high-resolution power information in networking, storage, memory and processing for primary embedded networked sensing applications. Using these results obtained in real-time we show that for a large class of wireless sensor network nodes, there exist several interdependencies in energy consumption between different subsystems. Through the use of our measurement tools we demonstrate that by carefully selecting the system operating points, energy savings of over 60% can be achieved while retaining system performance.

This paper describes a new embedded networked sensor platform architecture that combines hardware and software tools providing detailed, fine-grained real-time energy usage information. We introduce the LEAP2 platform, a qualitative step forward over the previously developed LEAP and other similar platforms. LEAP2 is based on a new low power ASIC system and generally applicable supporting architecture that provides unprecedented capabilities for directly observing energy usage of multiple subsystems in real-time. Real-time observation with microsecond-scale time resolution now enables direct accounting of energy dissipation for each computing task as well as for each hardware subsystem. This new hardware architecture is exploited with our new software tools, etop and endoscope. A series of experimental investigations provide high-resolution power information in networking, storage, memory and processing for primary embedded networked sensing applications. Using these results obtained in real-time we show that for a large class of wireless sensor network nodes, there exist several interdependencies in energy consumption between different subsystems. Through the use of our measurement tools we demonstrate that by carefully selecting the system operating points, energy savings of over 60% can be achieved while retaining system performance.