China
Africa
Explore chapters and articles related to this topic
Mobile-Relay Forwarding in Opportunistic Networks
Published in Mohamed Ibnkahla, Adaptation and Cross Layer Design in Wireless Networks, 2018
Giuseppe Anastasi, Marco Conti, Andrea Passarella, Luciana Pelusi
Since nodes are typically energy-constrained devices, a power management strategy is needed to save energy and increase nodes’ lifetime. In the context of opportunistic networking, the objective of power management is to minimize energy consumption while missing as few contacts as possible to achieve an adequate performance level in terms of message latency and delivery ratio. Ideally, the node should sleep for most of the time and wake up only when the MR is within its communication range. In practice, this is infeasible because the node is not able to know exactly when the next contact will occur, unless the MR mobility pattern is known in advance (predictable mobility). Thus, the MR and the nodes agree on a discovery protocol that allows a timely MR discovery to the node with minimum energy consumption. Obviously, the discovery protocol can be optimized based on the knowledge available about the MR mobility.
System-Level Power Management
Published in Louis Scheffer, Luciano Lavagno, Grant Martin, EDA for IC System Design, Verification, and Testing, 2018
Naehyuck Chang, Enrico Macii, Massimo Poncino, Vivek Tiwari
Dynamic voltage scaling is the power management technique that controls the supply voltage according to the current workload at run-time to minimize the energy consumption without having an adverse effect on system performance. Dynamic voltage scaling can be viewed as a variant of DPM in which DPM is applied not just to idle components but also to those resources that are noncritical in terms of performance, running the resource at different power/speed points. In other words, DVS introduces the notion of multiple active states, besides multiple idle states exploited by traditional DPM. Moreover, since in DVS power/speed trade-off points are defined by different supply voltage levels, DVS is traditionally applied to CPUs, rather than to other components, and it is thus exploited at the task granularity.
Customers: Electric Service Requirements
Published in J. Lawrence, P.E. Vogt, Electricity Pricing, 2017
Power management is the process of optimizing the utilization of electricity to extract efficiency while minimizing its cost. The prices of electric service may offer an impetus for controlling, where possible, the overall operation of load devices. If bill amounts are high, invoking changes in energy intensities or electricity usage patterns can prove to be economical even when capital expenditures may be necessary to achieve such operating versatility. On the other hand, if bill amounts are relatively low, the small potential for economical gains would not serve as a primary driver for customers to change their methods of energy utilization. Thus, changes in electricity usage depend upon a customer’s degree of operational and economic flexibility.8
Review on Multi-Port DC–DC Converters
Published in IETE Technical Review, 2021
Mudadla Dhananjaya, Swapnajit Pattnaik
A bidirectional port is designed for energy storage and two unidirectional input power ports are designed to interface the asymmetrical energy sources [48,49]. It can achieve better power management between the sources, battery, and load. Nevertheless, it increases the intricacy of the controller design which leads to, the system becoming complex. A systematic approach is introduced for developing the different MISO and SIMO topologies in [50]. It has single-stage power conversion between input sources and load, which results in high efficiency. However, it has more device count compared with recent, past proposed MISO, SIMO topologies which results, high cost, and increase the size of the converter. Boost converter based three-inputs/one output power electronic converter (MIPEC) is presented in [51], which is tested in high power rating (60 kW) for hybrid electric vehicle applications. It has achieved good dynamic behavior under several driving cycles.
Efficient resource management techniques in cloud computing environment: a review and discussion
Published in International Journal of Computers and Applications, 2019
Frederic Nzanywayingoma, Yang Yang
According to [39], Dynamic Component Deactivation (DCD), Dynamic Performance Scaling (DPS), and Dynamic Voltage and Frequency Scaling (DVFS) are ones of the dynamic power management techniques suggested. DPS technique is for the automatic adjustment of the performance proportional to the power consumption. DVFS and DCD are applied to different computer component and at OS level such as CPU, Memory, disk, network interface, other power-aware OS such as KVM, VMware solution, Xen Hypervisor. The application of DVFS technique decreases power consumption of a computing resource significantly. This technique was firstly used in portable and laptop systems to conserve battery power, and now it has been implemented on the server chipsets. Lowering the CPU frequency may lead to power savings and potential energy savings but may also impact application performance. Therefore, to maximize the energy efficiency while meeting the SLA constraints, scheduling algorithms are involved to determine a good operating frequency of the CPU to meet the application deadlines. The scheduling algorithms have to consider both cost and energy factors on the decision-making [40].
DT-MG: many-to-one matching game for tasks scheduling towards resources optimization in cloud computing
Published in International Journal of Computers and Applications, 2021
Yassir Samadi, Mostapha Zbakh, Claude Tadonki
We have compared our proposed algorithm (DT-MG) with six other algorithms, which are: Non-power aware (NPA): It does not apply any energy optimization. In this policy, all nodes operate at 100% CPU usage and consume maximum power all the time.Dynamic voltage and frequency scaling (DVFS): It is a commonly used power-management technique where the clock frequency of a processor is decreased to allow a corresponding reduction in the supply voltage. This reduces power consumption, which can lead to a significant reduction in the energy required to run a job [41].Static threshold and minimum migration time (ThrMmt): It uses a static Upper Threshold for detecting over-loaded nodes and Minimum Migration Time for selecting tasks to be migrated. [26].Median absolute deviation and minimum migration time (MadMmt): In this technique, the overload threshold is calculated dynamically using median absolute deviation and, then, it selects a task to migrate which has the least computational length as it will be migrated faster [42].Median absolute deviation and random selection (MadRs): This technique uses the median absolute deviation to detect the over-loaded PMs and migrates tasks randomly without applying any rules [26].Utilization and minimum correlation (UMC): It takes both machine utilization and machine correlation with VM for choosing a suitable machine to host the migrated tasks. They have also proposed VM-based dynamic threshold (VDT) algorithm for detecting under-loaded hosts and local regression (LR) technique for detecting the over-loaded hosts [43].Modified best fit decreasing (MBFD): It presents an energy-aware resource allocation policy for efficient datacenters management in cloud computing. It used a static upper and lower thresholds for detecting the over-loaded and under-loaded hosts. Then, a minimization of migration policy (MM) is used to migrate the minimum number of tasks. After that, a MBFD algorithm is applied to assign each migrated task to a host that gives the least increase of power consumption [19].