Pull-based overlays are used in some of today's largest computational grids. Job agents are submitted to resources with the duty of retrieving real workload from a central queue at runtime and executing it. This m...
详细信息
Pull-based overlays are used in some of today's largest computational grids. Job agents are submitted to resources with the duty of retrieving real workload from a central queue at runtime and executing it. This model helps overcome the problems of direct job submission in the highly complex grid environments, namely, heterogeneity, imprecise status information, relatively high failure rates and slow adaptation to changes of grid conditions or user priorities. This article presents a distributed scheduling architecture for such late-binding overlays. In this architecture, execution nodes share a distributed hash table and cooperatively perform job assignment. As our experiments prove, scalability problems of centralized matching are avoided, achieving low and predictable scheduling overheads even for execution of large workflows, and total turnaround times are improved. This is in line with the predictions of a theoretical model of grid workflow execution that the article also discusses. Scalability makes fine-grained scheduling possible and enables new functionalities, like a distributed data cache shared by the execution nodes, which helps alleviate the commonly congested storage services. In addition, we show that our system is more resilient to problems like communication breakdowns between computation centres. Moreover, the new architecture is better prepared to deal with demanding scenarios like intense demand of popular data files or remote data processing.
The proliferation of high powered electric devices is a driving force in the rising of peak power demand from electric power utilities. One way to accommodate these rising consumption patterns involves the deployment ...
详细信息
The proliferation of high powered electric devices is a driving force in the rising of peak power demand from electric power utilities. One way to accommodate these rising consumption patterns involves the deployment of high capacity dispatchable, but largely unsustainable peak generation systems. To avert these extravagant costs and the likelihood of grid overload, demand response (DR) strategies can be employed to curtail overall consumption, thus reducing peak patterns. In this paper, we propose a distributed real-time DR approach. The proposed method fosters seamless cooperation between DR participants for rapid convergence to expected aggregate load curtailment, while accounting for individual consumer satisfaction needs. We assess this paper through theoretical analysis based on population game theory and simulations to demonstrate its inherent flexibility, scalability, and resilience making it attractive for practical widespread deployment.
In this paper, we study the problem of real-time optimal distributed partitioning for perimeter patrolling in the context of multicamera networks for surveillance, where each camera has limited mobility range and spee...
详细信息
In this paper, we study the problem of real-time optimal distributed partitioning for perimeter patrolling in the context of multicamera networks for surveillance, where each camera has limited mobility range and speed, and the communication is unreliable. The objective is to coordinate the cameras in order to minimize the time elapsed between two different visits of each point of the perimeter. We address this problem by casting it into a convex problem in which the perimeter is partitioned into nonoverlapping segments, each patrolled by a camera that sweeps back and forth at the maximum speed. We then propose an asynchronous distributed algorithm that guarantees that these segments cover the whole patrolling perimeter at any time and asymptotically converge to the optimal centralized solution under reliable communication. We finally modify the proposed algorithm in order to attain the same convergence and covering properties even in the more challenging scenario, where communication is lossy and there is no channel feedback, i.e., the transmitting camera is not aware whether a packet has been received or not by its neighbors.
We describe a protocol for the average consensus problem on any fixed undirected graph whose convergence time scales linearly in the total number nodes n. The protocol relies only on nearest -neighbor interactions but...
详细信息
We describe a protocol for the average consensus problem on any fixed undirected graph whose convergence time scales linearly in the total number nodes n. The protocol relies only on nearest -neighbor interactions but requires all the nodes to know the same upper bound U on the total number of nodes which is correct within a constant multiplicative factor. As an application, we develop a distributed protocol for minimizing an average of (possibly nondifferentiable) convex functions (1/n) Sigma(n)(i=1) f(i)(theta) in the setting where only node i in an undirected, connected graph knows the function f(i)(theta). Under the same assumption about all nodes knowing U, and additionally assuming that the subgradients of each f(i)(theta) have absolute values bounded by some constant L known to the nodes, we show that after T iterations our protocol has error which is O(L root n/T). As a consequence, the time to solve this distributed optimization problem to any fixed accuracy is also linear in the number of nodes n.
Massive multiple-input multiple-output (mMIMO) is emerging as a cornerstone technology for fifth-generation (5G) communications. It promises to scale up the performance of the conventional communication systems by gro...
详细信息
Massive multiple-input multiple-output (mMIMO) is emerging as a cornerstone technology for fifth-generation (5G) communications. It promises to scale up the performance of the conventional communication systems by growing the number of antennas at the base station side. This paper proposes a decentralized, scalable, and energy-efficient radio resource allocation method tailored for the uplink of the upcoming 5G air interface, based on the mMIMO physical layer. The proposed solution elaborates on a game-theoretical approach, which aims at maximizing the energy efficiency of mobile terminals, while guaranteeing the respect of average data rates and power consumptions constraints. This formulation leads to a low-complexity, iterative, and distributed algorithm, which considers (just to mention few relevant issues) the impact of channel time selectivity, delayed feedback from the base station, and physical-layer details of the selected communication technology. An extensive simulation campaign, considering a long-term evolution-advanced-based multicellular system based on mMIMO, is used to evaluate the benefits of the proposed technique. By calculating energy efficiency, user and peak data rates, spectral efficiency, outage probability, and other minor performance indexes, the reported results clearly demonstrate the performance gain that the designed solution offers with respect to baseline strategies.
Current research in the field of distributed consensus algorithms fails to adequately address physical limitations of real systems. This paper proposes a new algorithm for quantized distributed load balancing over a n...
详细信息
Current research in the field of distributed consensus algorithms fails to adequately address physical limitations of real systems. This paper proposes a new algorithm for quantized distributed load balancing over a network of agents subject to upper-limit constraints. More precisely, loads are integer values, and nodes are constrained to remain under maximum load capacities at all times. Convergence to a set of balanced states is proven for all connected graphs with any feasible initial load distribution, given some conditions on the placement of nodes with finite capacity. We present simulations that verify our results and discuss possible extensions of the algorithm.
The diversity of components in the smart grid and issues, such as scalability, stability, and privacy, have led to the desire for more distributed control paradigms. In this paper, we address the problem of optimizing...
详细信息
The diversity of components in the smart grid and issues, such as scalability, stability, and privacy, have led to the desire for more distributed control paradigms. In this paper, we address the problem of optimizing smart grid operation with separable global costs and separable but nonconvex constraints, while considering important aspects of network operation, such as power flow and nodal voltage constraints. A localized primal dual method is applied through the use of an augmented Lagrange function, which is used to overcome the issues of non-convexity in the presence of nonlinear equality constraints. The nonseparability of the augmented Lagrange penalty function is addressed through the use of local and neighborhood communication leading to a completely distributed solution of the global problem.
This paper presents the first distributed algorithm to construct a spanner for arbitrary ad hoc networks under the physical signal-to-interference-and-noise-ratio (SINR) interference model. Spanner construction is one...
详细信息
This paper presents the first distributed algorithm to construct a spanner for arbitrary ad hoc networks under the physical signal-to-interference-and-noise-ratio (SINR) interference model. Spanner construction is one of the most important techniques for topology control in wireless networks, which intends to find a sparse topology in which only a small number of links need to be maintained, without substantially degrading the path connecting any pair of the nodes in the network. Due to the non-local property of interference, constructing a spanner is challenging under the SINR model, especially when a local distributed algorithm is desired. We meet this challenge by proposing an efficient randomized distributed algorithm that can construct a spanner in O(log n log Gamma) timeslots with a high probability, where n is the total number of nodes and Gamma describes the ratio of the maximum distance to the minimum distance between nodes. The constructed spanner concurrently satisfies two most desirable properties: constant stretch and linear sparseness. Our algorithm employs a novel maximal independent set (MIS) procedure as a subroutine, which is crucial in achieving the time efficiency of spanner construction. The MIS algorithm improves the best known result of O(log(2) n) [33] to O(log n) and is of independent interest as the algorithm is applicable also to many other applications. We conduct simulations to verify the proposed spanner construction algorithm, and the results show that our algorithm also performs well in realistic environments.
This study proposes a distributed outlier detection algorithm based on credibility feedback in wireless sensor networks. The algorithm consists of three stages, which are evaluating the initial credibility of sensor n...
详细信息
This study proposes a distributed outlier detection algorithm based on credibility feedback in wireless sensor networks. The algorithm consists of three stages, which are evaluating the initial credibility of sensor nodes, evaluating the final credibility based on credibility feedback and Bayesian theorem, and adjusting for the outlier set. Simulation results verify that the algorithm can achieve high detection accuracy and low false alarm rate, even in the situation that the network with a large number of outliers.
We consider the problem of distributed learning, where a network of agents collectively aim to agree on a hypothesis that best explains a set of distributed observations of conditionally independent random processes. ...
详细信息
We consider the problem of distributed learning, where a network of agents collectively aim to agree on a hypothesis that best explains a set of distributed observations of conditionally independent random processes. We propose a distributed algorithm and establish consistency, as well as a nonasymptotic, explicit, and geometric convergence rate for the concentration of the beliefs around the set of optimal hypotheses. Additionally, if the agents interact over static networks, we provide an improved learning protocol with better scalability with respect to the number of nodes in the network.
暂无评论