Loading...
Search for:
energy-efficiency
0.008 seconds
Total 357 records
Improving the Utilization Rate of the GPU Memory Resources by Exploiting Application's Heterogeneity
, Ph.D. Dissertation Sharif University of Technology ; Sarbazi Azad, Hamid (Supervisor)
Abstract
In recent years, GPUs have become a popular choice for high-performance general-purpose systems. However, since general-purpose applications do not always utilize processing and memory resources efficiently, energy consumption in these systems has not been managed effectively. As the variety of resources within GPUs continues to increase, the problem of low utilization has become more critical. Past solutions have focused on integrating resources and using multi-tasking, but these methods have limitations in terms of performance and security. Therefore, this research proposes new methods for improving the resource utilization and energy efficiency of GPUs. This dissertation first examines...
An Approach for Dynamic Selection of Virtual Machines for Migration in Cloud Data Centers
, M.Sc. Thesis Sharif University of Technology ; Movaghar Rahimabadi, Ali (Supervisor)
Abstract
Increasing the workload of physical machines in data centers, performance of the system decreases and energy consumption increases. Various methods have been proposed to overcome this challenge and solve the problem. One of these solutions is the migration of virtual machines (VMs) from the over-loaded physical machine (PM) to another one, namely the under-loaded PM, in the clouds, which is effective in increasing the performance of the system and improving energy consumption When the performance of a PM is less than a certain threshold, due to overload, migration of the selected VM to a candidate PM can improve performance, if it is done in an intelligent way. Choosing the best suitable VM...
A Real-Time and Energy-Efficient Decision Making Framework for Computation Offloading in Iot
, M.Sc. Thesis Sharif University of Technology ; Fazli, Mohammad Amin (Supervisor)
Abstract
Based on fog computing paradigm, new applications have become feasible through the use of hardware capabilities of smart phones. Many of these applications require a vast amount of computing and real-time execution should be guaranteed. Based on fog computing, in order to solve these problems in is necessary to offload heavy computing to servers with adequate hardware capabilities. On the other side, the offloading process causes time overhead and endangers the real-timeliness of the application. Also, because of the limited battery capacity of the handheld devices, energy consumption is very important and should be minimized.The usual proposed solution for this problem is to refactor the...
An Efficient Reconfigurable Architecture to Speed up Machine Learning Algorithms
, M.Sc. Thesis Sharif University of Technology ; Asadi, Hossein (Supervisor)
Abstract
Machine learning algorithms are achieving state-of-the-art performance in many various applications such as image processing, machine vision, speech recognition, diagnosis diseases, robotics, military, and aerospace. For decades, the usage of machine learning algorithms especially Neural Network Algorithms (NNA) has been restricted due to their complexity and high computation time of available inefficient hardwares. Although advances in technology and the emergence of powerful processors has increased the usage of NNAs, especially Deep Neuaral Networks (DNNs), the research gap in machine learning hardware platform with high performance as well as high energy efficiency is still remaining....
Energy Optimization in Roadside Unit in Vehicular Ad-hoc Networks
, M.Sc. Thesis Sharif University of Technology ; Afshin Hemmatyar, Ali Mohammad (Supervisor)
Abstract
Optimization of energy consumption is generally important in mobile networks and in particular in automotive networks. One of the challenges of installing roadside units in automotive case networks is to provide energy to the equipment assigned to these units. In some cases, power is easily available, but in many places, the cost of energy is very high and does not have economic justification. Therefore, optimization of energy consumption in roadside unit is of great importance. Typically, vehicles limitations in terms of energy consumption for communications are far lower than that of a roadside unit. In this dissertation, part of roadside communication to vehicles is shifted to the...
Energy Efficient Software Defined Networks
, M.Sc. Thesis Sharif University of Technology ; Hemmatyar, Ali Mohammad Afshin (Supervisor)
Abstract
With a growing concern on the considerable energy consumed by data centers, research efforts are targeting toward green data centers with higher energy efficiency. Networks are essential to today’s highly instrumented and connected society. Currently responsible for a fraction of the energy the IT sector consumes, networks rely on equipments that often use a constant amount of power regardless of their utilization; a worsening scenario as traffic is expected to increase by a factor of three in the next few years. Datacenter networks, for instance, employ devices that consume several kilowatts of power even when idle. This work aims to improve the energy efficiency of datacenter networks by...
Improving Energy Efficiency in Multi-processor Soft-Core Systems Using System-level Techniques
, M.Sc. Thesis Sharif University of Technology ; Goudarzi, Maziar (Supervisor)
Abstract
The ever increasing density and performance of FPGAs, has increased the importance and popularity of soft processors. One major research concern in this regard lays in the field of energy efficiency of the system on FPGA. This work is particularly focused on the energy efficiency of multiprocessor structures on FPGA using system level techniques. The growing gap between the speed of processors and memories can partly be compensated through memory hierarchy, i.e. caches. Since memory accesses follow a non-uniform distribution, and vary from one application to another, variable set-associative cache architectures have emerged. In this thesis, two novel cache architecture, primarily aimed at...
Application Communication-aware Data Center Energy Optimization
, M.Sc. Thesis Sharif University of Technology ; Goudarzi, Maziar (Supervisor)
Abstract
Virtualization has provided good and efficient use of resources in today datacenters. In this environment, modules of multi-tier applications can run on several Virtual Machines (VMs). VMs that are dedicated to a module often need to transfer data among themselves, and also to the VMs in other modules, and hence, the overall application runtime may be increased if the communicating VMs are put on far physical machines; consequently, servers must be up for a longer time and will consume more energy. Consolidating several VMs on a single Physical Machine (PM) effectively reduces the total computing power consumption, but we show that it is also important to consider the above inter-VM...
Heat and Mass Transfer in a Finite Channel
, M.Sc. Thesis Sharif University of Technology ; Saeedi, Mohammad Hassan (Supervisor)
Abstract
The need for cooling and creating thermal comfort conditions for human life is one of today's severe challenges. Therefore, development and innovation to enhance the efficiency of these systems are of great importance. Among the new ideas proposed in this domain, we can mention the M-cycle, which can greatly improve the energy consumption for cooling systems in this sector. In this cycle, the desired cooling is produced by utilizing the latent heat of water evaporation, which is considered a kind of renewable energy. The systems based on this cycle are still under development, and for this reason, there is a necessity to better investigate the heat and mass transfer in the aforementioned...
Enhancing Energy Efficiency in Ideal Binary Distillation tower with a Dynamic Process Intensification
, M.Sc. Thesis Sharif University of Technology ; Vafa, Ehsan (Supervisor)
Abstract
In this thesis titled "Enhancing Energy Efficiency in Ideal Binary Distillation tower with a Dynamic Process Intensification," operational methods in distillation units are explored and improved. Given the high cost of energy, limited energy resources, and environmental considerations, increasing energy efficiency in process industries has become significantly important. In this context, this research attempts to enhance the performance of the distillation unit by utilizing dynamic process intensification with two operational points corresponding to the production of products with purities both higher and lower than the desired product purity, and switching between these two points. The...
Management of Cryptocurrency Mining Penetration in Electricity Distribution Networks and Provision of the Solutions
, Ph.D. Dissertation Sharif University of Technology ; Vakilian, Mehdi (Supervisor) ; Hajipour, Ehsan (Supervisor)
Abstract
Currently, cryptocurrencies have attained a significant role in the financial transactions of the world, especially due to their peer-to-peer and decentralized characteristics, resulting in the elimination of intermediary oversight entities. In the cryptocurrency network, transaction security is ensured through devices known as cryptocurrency miners. People place their cryptocurrency mining devices at the disposal of the cryptocurrency network and, depending on their device’s processing power and capacity, receive rewards in the form of cryptocurrencies. Given the complexity of cryptocurrency networks, regular devices lack the ability to mine cryptocurrencies, necessitating high processing...
Deep Reinforcement Learning for Building Climate Control Using Weather Forecast Data
, M.Sc. Thesis Sharif University of Technology ; Rezaeizadeh, Amin (Supervisor)
Abstract
Buildings account for more than 30% of the world’s total energy consumption. Among building end-uses, air conditioning and in particular cooling systems have a major share of more than 50%. Therefore, design of optimal controllers for AC systems has become increasingly important. Classical and model-free control methods typically lack the ability to optimize energy consumption. On the other hand, model-based optimal control methods rely on precise modeling, which is difficult to acquire due to the complexity of the AC system dynamics.In recent years, deep reinforcement learning has become a popular choice for optimal control of systems with complex dynamics. In this thesis, a deep...
Resource Management in Space-Division Multiplexed Elastic Optical Networks
, M.Sc. Thesis Sharif University of Technology ; Pakravan, Mohammad Reza (Supervisor)
Abstract
In recent years, Internet traffic has nearly doubled in the core network every two years, and it is expected to continue to grow in the future. With this growth, there is a need for very high capacity and economically efficient transmission systems, which optical communication is the best choice. Elastic optical network (EON) is one of the best options for achieving this goal in the future. Unlike existing networks, the transmission parameters of EON will be customizable, which will have an essential role in optimizing bandwidth assignment. The problem facing scientists today is to approach the final physical capacity of current optical fibers. The solution proposed for this challenge is...
Design of an Energy Efficient Routing Algorithm for WSNs
, M.Sc. Thesis Sharif University of Technology ; Pakravan, Mohammad Reza (Supervisor)
Abstract
Wireless Sensor Networks (WSNs) have developed in recent years and are used in many different applications such as fire detection, military operations, and healthcare. Sensor nodes can sense an environment, sending and receiving information and forwarding packets from source nodes to a sink. These tasks consume large amounts of energy. In many WSN applications, delay and reliability of data delivery and the energy consumption of nodes are important design metrics. Network protocols such as medium access control (MAC) and routing should be designed to improve these metrics In this thesis, we propose a new cross layer algorithm that improves the performance of WSN networks based on these...
Analysis and Improvement of Energy-Efficient Routing Algorithms for Delay Tolerant Networks
, M.Sc. Thesis Sharif University of Technology ; Pakravan, Mohammad Reza (Supervisor)
Abstract
Delay Tolerant Networks (DTNs) are sparse networks in which intermittent connectivity happens due to energy shortage or high mobility and long latency is unavoidable.These networks have a wide range of applications in our daily life such as internet of moving things. In delay tolerant networks, finding an end to end path between source and destination is almost impossible. Therefore, routing can’t be done based on finding an end to end path but it has to be done based on hop by hop sessions. In this thesis, we aim to represent an efficient routing algorithm for delay tolerant networks that can improve delivery probability, delivery cost and delivery latency.Our proposed routing algorithm is...
Long Term Impacts of Electricity Pricing Methods on Multi-Carrier Energy Systems with High PHEV Penetration
, M.Sc. Thesis Sharif University of Technology ; Ranjbar, Ali Mohammad (Supervisor)
Abstract
As far as reliable supply and cost-effective methods of power delivery to end-users are concerned, new advancements in smart grids and, in particular, smart pricing are of high significance to grid operators. In the recent decade, there has been a considerable raise in penetration of renewable energy resources and also the number of hybrid electric vehicles which can act as a huge mobile storage in the network. Hence, finding a practical and cost-effective method of management of all these new resources and avoiding potential problems associated with them seems to be of high importance. In this study, first, the response of demand to pricing signals has been modeledAfterwards, the response...
Appropriate MAC Sub-layer Algorithms for Body Area Networks and Related Analysis
, M.Sc. Thesis Sharif University of Technology ; Nasiri-Kenari, Masoumeh (Supervisor) ; Vosougi Vahdat, Bijan (Co-Advisor)
Abstract
WBAN , as one of the most important wireless networks in the near future, has received significant attentions in recent years. The most important designing challenge in these networks is the scarcity of energy resources especially in their sensor nodes. Additionally, special characteristics of these networks have led the other network’s designs, especially the general sensor network protocols to be inefficient for the WBAN. Among others, the existences of recurring patterns in each sensor channel fading and the high correlation among different sensor channels are some of the most important characteristics of WBANs. Considering these characteristics, in this thesis, the MAC sub-layer of...
Improving direct load control implementation by an inititative load control method
, Article EPDC 2013 - 18th Electric Power Distribution Network Conference ; 2013 ; ISBN: 9781480000000 ; Fotuhi-Firuzabad, M ; Moeini-Aghtaie, M ; Sharif University of Technology
Abstract
Demand response (DR) as a state-of-the-art program is accommodated in the energy efficiency contexts of the smart grid. Lack of the customer knowledge about how to respond to the time-differentiated tariffs and offered controlling signals is the major obstacle in the way of broad DR implementing. As a solution, an optimization method, namely load control (LC), is proposed to automatically control the on/off status of the responsive appliances in a smart home. This paper focuses on the direct load control (DLC) program in the category of incentive-based DR programs. LC is extended to consider inconvenience cost of the DLC implementation for DLC program participants. Outputs of the extended LC...
Impacts of plug-in hybrid electric vehicle uncertainty and grid unavailability on home load management
, Article 2012 11th International Conference on Environment and Electrical Engineering, EEEIC 2012 - Conference Proceedings ; 2012 , p. 693-698 ; ISBN: 9781460000000 ; Safdarian, A ; Fotuhi-Firuzabad, M ; Aminifar, F ; Sharif University of Technology
Abstract
A massive focus has recently been made on demand response (DR) programs, aimed to the electricity price reduction, reliability improvement, and energy efficiency. Basically, DR programs are divided into twofold main categories, namely incentive-based programs and price- or time-based programs. The focus of this paper is on priced-based DR programs including consumer responses to the time differentiated pricing. Home load management (HLM) program is designed to control responsive appliances and charging/discharging cycles of plug-in hybrid electric vehicles (PHEVs) by the consumer. Uncertain parameters associated with PHEV, i.e. its departure/travelling time and energy consumption as well as...
Power-efficient prefetching on GPGPUs
, Article Journal of Supercomputing ; Volume 71, Issue 8 , August , 2015 , pp. 2808-2829 ; ISSN: 09208542 ; Hessabi, S ; Abdi, M ; Baniasadi, A ; Sharif University of Technology
Abstract
The graphics processing unit (GPU) is the most promising candidate platform for achieving faster improvements in peak processing speed, low latency and high performance. The highly programmable and multithreaded nature of GPUs makes them a remarkable candidate for general purpose computing. However, supporting non-graphics computing on graphics processors requires addressing several architectural challenges. In this paper, we focus on improving performance by better hiding long waiting time for transferring data from the slow global memory. Furthermore, we show that the proposed method can reduce power and energy. Reduction in access time to off-chip data has a noticeable role in reducing...