Aeolus is a prototype implementation of a topology optimizer on top of the distributed streaming system Storm. Aeolus extends Storm with a batching layer which can increase the topology's throughput by more than o...
详细信息
ISBN:
(纸本)9781467349093;9781467349086
Aeolus is a prototype implementation of a topology optimizer on top of the distributed streaming system Storm. Aeolus extends Storm with a batching layer which can increase the topology's throughput by more than one order of magnitude. Furthermore, Aeolus implements an optimization algorithm that computes the optimal batch size and degree of parallelism for each node in the topology automatically. Even if Aeolus is built on top of Storm, the developed concepts are not limited to Storm and can be applied to any distributed intra-node-parallel streaming system. We propose to demo Aeolus using an interactive Web UI. One part of the Web UI is a topology builder allowing the user to interact with the system. Topologies can be created from scratch and their structure and/or parameters can be modified. Furthermore, the user is able to observe the impact of the changes on the optimization decisions and runtime behavior. Additionally, the Web UI gives a deep insight in the optimization process by visualizing it. The user can interactively step through the optimization process while the UI shows the optimizer's state, computations, and decisions. The Web UI is also able to monitor the execution of a non-optimized and optimized topology simultaneously showing the advantage of using Aeolus.
The Branch-and-Bound (B& B) method is a wellknown optimization algorithm for solving integer linear programming (ILP) models in the field of operations research. It is part of software often employed by businesses...
详细信息
ISBN:
(纸本)9781479907298
The Branch-and-Bound (B& B) method is a wellknown optimization algorithm for solving integer linear programming (ILP) models in the field of operations research. It is part of software often employed by businesses for finding solutions to problems such as airline scheduling problems. It operates according to a divide-and-conquer principle by building a treelike structure with nodes that represent linear programming (LP) problems. A LP solver commonly used to process the nodes is the simplex method. Nowadays its sequential implementation can be found in almost all commercial ILP solvers. In this paper, we present a hybrid CPU-GPU implementation of the B&B algorithm. The B&B tree is managed by the CPU, while the revised simplex method is mainly a GPU implementation, relying on the CUDA technology of NVIDIA. The CPU manages concurrently multiple instances of the LP solver. The principal difference with a sequential implementation of the B&B algorithm pertains to the LP solver, provided that the B&B tree is managed with the same strategy. We thus compared our GPU-based implementation of the revised simplex to a wellknown open-source sequential solver, named CLP, of the COINOR project. For given problem densities, we measured a size threshhold beyond which our GPU implementation outperformed its sequential counterpart.
Private Mobile Radio (PMR) networks are cellular infrastructures dedicated to be used by professionals, such as public safety, military, industry and transportation organizations. In those networks, resources are scar...
详细信息
ISBN:
(纸本)9781467359399;9781467359382
Private Mobile Radio (PMR) networks are cellular infrastructures dedicated to be used by professionals, such as public safety, military, industry and transportation organizations. In those networks, resources are scarce, and there are strong Quality of Service (QoS) requirements. The emergence of new services which need more bandwidth has made the world PMR leader focus on the LTE-Advanced protocol. In order to ensure the QoS despite resource shortage, we propose an algorithm of Resource Blocks (RBs) allocation with spatial frequency reuse whose scheme takes into account users' (UEs') interference possibility and probability. We start by defining the underlying problem, which we call Weighted Fractional Coloring Problem (WFCP), in terms of graph theory. Next, we prove its NP-hardness. As obtaining an exact solution of such a problem in reasonable time is unrealistic, we propose a heuristic algorithm. In order to evaluate the performance of our algorithm we use a rigorous validation procedure. We compare its performance with that of a random one which we propose as a reference and the exact one which can be run on very small networks. Thanks to the results obtained we believe that the proposed algorithm can establish a solid starting point to conceive its distributed versions for novel PMR protocols.
Along with all the positives brought by the Internet, the global network is also used for criminal purposes. The goal of the presented work was to create and optimize applications working in parallel with the search a...
详细信息
ISBN:
(纸本)9783642385599
Along with all the positives brought by the Internet, the global network is also used for criminal purposes. The goal of the presented work was to create and optimize applications working in parallel with the search and illegal content analysis system (SWAT) created previously at the authors' university. The role of the SWAT system is to penetrate the Internet to search for images and provide links to the presented application. The presented application is able to detect various special objects, such as faces, symbols or child pornography. The use of bees algorithm-based optimization made it possible to increase the effectiveness of image analysis. With appropriate assumptions, the efficiency was increased by a factor of three times compared to the application without optimization. The efficiency of the optimization algorithm depends largely on the nature of the input data, the URL addresses.
The key issue of the localization study is that how we can minimize the energy consumption of devices with guaranteeing high degree of accuracy. In this paper, we show that the collaboration among proxy devices with s...
详细信息
ISBN:
(纸本)9781612848242
The key issue of the localization study is that how we can minimize the energy consumption of devices with guaranteeing high degree of accuracy. In this paper, we show that the collaboration among proxy devices with short range communication is helpful to energy-efficiently localize their locations in time-average sense by analyzing the device proximity including real GPS trace of students in KAIST and NCSU campuses. Next, we deliberate what is the best method for selfish mobile users to collaborate for the energy-efficient localization, and formulate an optimization problem which considers the energy efficiency and/or user fairness. However, optimizing this problem is tricky since it requires a global knowledge of sets of proxy devices and also solving a NP-hard problem to select devices which directly measure locations. This paper makes a contribution towards presenting a practical and fully distributed location sharing protocol based on competition for turning off GPS, and an optimal algorithm which controls mean waiting time used for the competition. Through the extensive simulations under several sample topologies and real mobility trace in KAIST campus, we obtain the following interesting observations: (i) (in sample topologies) our scheme achieves a near-optimal performance of proposed problem in terms of energy efficiency and fairness (up to 27.2% power saving with 35.8% higher fairness than existing heuristic algorithms), (ii) (in real mobility trace) our scheme well adapts at even unpredictably changing mobility environment (65.5% power saving than no collaboration, 27.4% or more power saving with 25% higher fairness than the existing algorithms).
In this paper, we present a novel linear model predictive control (MPC) scheme that relies on a continuous-time, barrier function based algorithm which asymptotically tracks the solution of a time-varying open-loop op...
详细信息
ISBN:
(纸本)9783033039629
In this paper, we present a novel linear model predictive control (MPC) scheme that relies on a continuous-time, barrier function based algorithm which asymptotically tracks the solution of a time-varying open-loop optimal control problem. In particular, the control input is obtained as the sampled output of a continuous-time dynamical system and no iterative optimization algorithm is needed in the on-line implementation. In addition, we present a new approach towards stabilizing MPC based on gradient recentered logarithmic barrier functions that allows enlargement of the employed terminal set.
Reducing fluctuations and mitigating the variability of wind energy is essential for its integration in grids, electricity markets, microgrids, and distributed generation settings. In this work, we present a storage-b...
详细信息
ISBN:
(纸本)9781467353304;9781467353298
Reducing fluctuations and mitigating the variability of wind energy is essential for its integration in grids, electricity markets, microgrids, and distributed generation settings. In this work, we present a storage-based wind power smoothing system that uses novel optimization algorithms to reduce the variability of wind energy. The system considers forecasted and actual energy generated, battery size, and energy prices and determines export rates that have low variability and maximize either the energy exported or revenue earned. Our optimization algorithms are novel as they model an equivalent relaxed buffer system that uses only linear constraints and allows the computation of optimal smoothing solutions in an efficient manner. This enables the system to be used in an online manner in real time as well as in planning and operations. The smoothing system and the mathematical models and programs used in optimization are presented along with preliminary simulation results that demonstrate the need and effectiveness of the system with the help of real wind energy data. Finally, we compare wind smoothing to video smoothing and point out the important similarities and differences.
After a generation of writing and improving lens design software, it is time to assess where we are. Specifically, can a modern program compete with, or surpass, the best human designers? Here we describe a friendly c...
详细信息
ISBN:
(纸本)9780819496911
After a generation of writing and improving lens design software, it is time to assess where we are. Specifically, can a modern program compete with, or surpass, the best human designers? Here we describe a friendly contest between two leaders in the field.
In cloud systems, it is non-trivial to optimize task's execution performance under user's affordable budget, especially with possible workload prediction errors. Based on an optimal algorithm that can minimize...
详细信息
ISBN:
(纸本)9781479907298
In cloud systems, it is non-trivial to optimize task's execution performance under user's affordable budget, especially with possible workload prediction errors. Based on an optimal algorithm that can minimize cloud task's execution length with predicted workload and budget, we theoretically derive the upper bound of the task execution length by taking into account the possible workload prediction errors. With such a state-of-the-art bound, the worst-case performance of a task execution with a certain workload prediction errors is predictable. On the other hand, we build a close-to-practice cloud prototype over a real cluster environment deployed with 56 virtual machines, and evaluate our solution with different resource contention degrees. Experiments show that task execution lengths under our solution with estimates of worstcase performance are close to their theoretical ideal values, in both non-competitive situation with adequate resources and the competitive situation with a certain limited available resources. We also observe a fair treatment on the resource allocation among all tasks.
Nowadays an increasing number of papers appears in the subject of combinatorial optimization proposing a great variety of heuristics and metaheuristics, most of them apply special solution to fit the particular proble...
详细信息
ISBN:
(纸本)9781424414406
Nowadays an increasing number of papers appears in the subject of combinatorial optimization proposing a great variety of heuristics and metaheuristics, most of them apply special solution to fit the particular problem type. The analysis points out the importance of the generalization but the special intelligence in the algorithm design is still very important. Although the navigation in the solution space can be realized implicitly it has a decisive role in the performance. It is important to note that the success of sophisticated methods is often reduced by their relatively bad explorative capability. The well known algorithms are compared and also the latest most successful methods based on fast and simple heuristics- are discussed.
暂无评论