Houssem R. E. H. Bouchekara (17743281), Abdulazeez F. Salami (17743284), Yusuf A. Sha’aban (16052906) et al. · Figshare · 2024
<div><p>This paper is a follow-up to a recent work by the authors on recoverable UAV-based energy-efficient reconfigurable routing (RUBER) scheme for addressing sensor node and route failure issues in smart wireless livestock sensor networks. Time complexity and processing cost issues connected to the RUBER scheme are consequently treated in this article by proffering a time-aware UAV-based energy-efficient reconfigurable routing (TUBER) scheme. TUBER scheme employs a synchronized clustering-with-backup strategy, a minimum-hop neighborhood recovery mechanism, and a redundancy minimization technique. Comparative network performance of TUBER was investigated and evaluated with respect to RUBER and UAV-based energy-efficient reconfigurable routing (UBER) schemes. The metrics adopted for this comparative performance analysis are Cluster Survival Ratio (CSR), Network Stability (NST), Energy Dissipation Ratio (EDR), Network Coverage (COV), Packet Delivery Ratio (PDR), Fault Tolerance Index (FTI), Load Balancing Ratio (LBR), Routing Overhead (ROH), Average Routing Delay (ARD), Failure Detection Ratio (FDR), and Failure Recovery Ratio (FRR). With reference to best-obtained values, TUBER demonstrated improvements of 36.25%, 24.81%, 34.53%, 15.65%, 38.32%, 61.07%, 31.66%, 63.20%, 68.96%, 66.19%, and 78.63% over RUBER and UBER in terms of CSR, NST, EDR, COV, PDR, FTI, LBR, ROH, ARD, FDR, and FRR, respectively. These experimental results confirmed the relative effectiveness of TUBER against the compared routing schemes.</p></div>
Cited by 0Open access
Yosi Ben-Asher · Undergraduate topics in computer science · 2012
Cited by 0
Peng Zhang (2071), Jialiang Zhang (1574704), Yi Li (1144) · Figshare · 2025
<div> Timely prediction of memory failures is crucial for the stable operation of data centers. However, existing methods often rely on a single classifier, which can lead to inaccurate or unstable predictions. To address this, we propose a new ensemble model for predicting CE-driven memory failures, where failures occur due to a surge of correctable errors (CEs) in memory, causing server downtime. Our model combines several strong-performing classifiers, such as Random Forest, LightGBM, and XGBoost, and assigns different weights to each based on its performance. By optimizing the decision-making process, the model improves prediction accuracy. We validate the model using in-memory data from Alibaba’s data center, and the results show an accuracy of over 84%, outperforming existing single and dual-classifier models, further confirming its excellent predictive performance. </div>
Cited by 0Open access
Allen D. Malony, Sameer Shende · Lecture notes in computer science · 2004
Cited by 22
Louis Breit · 2003
Cited by 2
Peng Zhang (2071), Jialiang Zhang (1574704), Yi Li (1144) · Figshare · 2025
<div> Timely prediction of memory failures is crucial for the stable operation of data centers. However, existing methods often rely on a single classifier, which can lead to inaccurate or unstable predictions. To address this, we propose a new ensemble model for predicting CE-driven memory failures, where failures occur due to a surge of correctable errors (CEs) in memory, causing server downtime. Our model combines several strong-performing classifiers, such as Random Forest, LightGBM, and XGBoost, and assigns different weights to each based on its performance. By optimizing the decision-making process, the model improves prediction accuracy. We validate the model using in-memory data from Alibaba’s data center, and the results show an accuracy of over 84%, outperforming existing single and dual-classifier models, further confirming its excellent predictive performance. </div>
Cited by 0Open access
Dušan Okanović, Milan Vidaković, Zora Konjović · 2013
Continuous monitoring tools are designed to perform well under production workload with minimal performance overhead. Standard AOP frameworks are popular choice for development of instrumentation for Java software monitoring tools. Inevitable consequence of using these tools is the occurrence of performance overhead. The code they generate is guaranteed to be correct, but always has a slight overhead. On the other hand, if a low-level bytecode manipulation tools are chosen, they may have better performance, but development time is higher and code may not be correct. The paper demonstrates the use of a new framework that allows instrumentation of Java code using AOP-like syntax, but has less overhead than industry standard AOP tools. Already developed monitoring tool has been implemented using this new framework, with better results.
Cited by 1
Heinemeyer, Christoph, Langosch, Katharina, Feldmann, Markus · RWTH Publications (RWTH Aachen) · 2010
Cited by 0
Michael Kurdziel, Marcin Łukowiak, Michael A. Sanfilippo · Journal of Cryptographic Engineering · 2013
Cited by 3
Elvin Abdullayev Elvin Abdullayev · PAHTEI-Procedings of Azerbaijan High Technical Educational Institutions · 2022
In recent years sales of cellular connection has become the fastest growing industry and one of the major areas of domestic and global telecommunications markets. The rapid growth of information flows requires accelerated development of the cellular market and continuous improvement methodology, economic theory, and methods of analysis and management in this area. The most important tasks of modern management practices are to develop and implement decisions to achieve financial and economic stability and efficiency of the organization.In this regard, there are problems with the need to develop, apply, and perspective development of theoretical-methodological and administrative - actual methods to cost analysis and product costs in the economic activities of mobile companies, which determines the relevance of the research topic. All this leads to problems associated with the need to develop theoretical and methodological, organizational, and practical approaches of management analysis of cellular companies in the modern economy of our country and determines the relevance of the research topic. From a cost management standpoint, the research suggests that manufacturing organizations employ the ABC method since it allocates overhead expenses more correctly than alternative cost accounting systems. Adopting this approach will result in: a) a better knowledge of the business processes that drive costs, b) a better technique for identifying productivity improvements that improve service delivery, and c) a clear identification of the performance metrics that improve cost management. These are critical elements of continuous improvement and comprehensive quality management programs that are critical to achieving an organization's objectives, goals, and, ultimately, financial success. Keywords: over head, costs, economic development
Cited by 0Open access
Mastilak, Matthew Christian · Michigan State University Libraries · 2007
Cited by 0Open access
Leonard J. Cimini, John MacLaren Walsh, Steven Weber et al. · 2015
Abstract : This project studied the tradeoff between the performance of resource and link controllers in distributed wireless networks and the amount of overhead measurement and control information utilized by the controllers. The key novel idea guiding the research was to view the control signals as messages in a distributed lossy source code, and the overhead performance tradeoff as a rate distortion curve. The project began by motivating the importance of the problem by authoring a thorough review of resource control signaling in the 4G cellular standards, which observed that roughly a quarter to a third of all downlink time-frequency resources are spent on non-information bearing control information that is not efficiently encoded. Three simple resource controllers were then investigated, and their overhead performance rate distortion function was calculated using a novel adaptation of the Blahut Arimoto algorithm to the CEO problem with independent sources. Practical source compressors and quantizers for both interactive and non-interactive control scenarios were developed that approached the associated fundamental limits. Extensions to controllers for multiuser MIMO communications and joint source channel codes for the control information were also considered.
Cited by 0
Young Yoon (3159627), Beom Heyn Kim (3159624) · Figshare · 2016
Cited by 0Open access
Jiyuan Zhang, Franz Franchetti, Tze Meng Low · arXiv (Cornell University) · 2018
The computation of convolution layers in deep neural networks typically rely on high performance routines that trade space for time by using additional memory (either for packing purposes or required as part of the algorithm) to improve performance. The problems with such an approach are two-fold. First, these routines incur additional memory overhead which reduces the overall size of the network that can fit on embedded devices with limited memory capacity. Second, these high performance routines were not optimized for performing convolution, which means that the performance obtained is usually less than conventionally expected. In this paper, we demonstrate that direct convolution, when implemented correctly, eliminates all memory overhead, and yields performance that is between 10% to 400% times better than existing high performance implementations of convolution layers on conventional and embedded CPU architectures. We also show that a high performance direct convolution exhibits better scaling performance, i.e. suffers less performance drop, when increasing the number of threads.
Cited by 32Open access
Matthew Sottile, Alan Morris, Allen Malony et al. · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 2007
Abstract. Online application performance monitoring allows tracking performance characteristics during execution as opposed to doing so post-mortem. This opens up several possibilities otherwise unavailable such as real-time visualization and application performance steering that can be useful in the context of long-running applications. As HPC sys-tems grow in size and complexity, the key challenge is to keep the online performance monitor scalable and low overhead while still providing a useful performance reporting capability. Two fundamental components that constitute such a performance monitor are the measurement and transport systems. We adapt and combine two existing, mature systems- TAU and Supermon- to address this problem. TAU performs the mea-surement while Supermon is used to collect the distributed measurement state. Our experiments show that this novel approach leads to very low-overhead application monitoring as well as other benefits unavailable from using a transport such as NFS.
Cited by 0Open access
Lakshmi Prasad Rongali · 2025
As distributed systems become more widespread, the performance overhead of observability tools like OpenTelemetry becomes a critical concern. This paper presents a systematic analysis of OpenTelemetry's performance impact, examining how different prog
Cited by 0Open access
Jie Ren, Bradford D. Boyle, Gwanmo Ku et al. · arXiv (Cornell University) · 2014
A key aspect of many resource allocation problems is the need for the resource controller to compute a function, such as the max or arg max, of the competing users metrics. Information must be exchanged between the competing users and the resource controller in order for this function to be computed. In many practical resource controllers the competing users' metrics are communicated to the resource controller, which then computes the desired extremization function. However, in this paper it is shown that information rate savings can be obtained by recognizing that controller only needs to determine the result of this extremization function. If the extremization function is to be computed losslessly, the rate savings are shown in most cases to be at most 2 bits independent of the number of competing users. Motivated by the small savings in the lossless case, simple achievable schemes for both the lossy and interactive variants of this problem are considered. It is shown that both of these approaches have the potential to realize large rate savings, especially in the case where the number of competing users is large. For the lossy variant, it is shown that the proposed simple achievable schemes are in fact close to the fundamental limit given by the rate distortion function.
Cited by 1Open access
Mastilak, Matthew Christian · Michigan State University Libraries · 2024
Cited by 0Open access
Ang, Dominic · DR-NTU (Nanyang Technological University) · 2021
Postactivation performance enhancement (PAPE) is a training method to maximise an athlete’s rate of force development (power output) with an acute improvement effect 1. There is currently a paucity of research in water-based and/or overhead throwing-related sports (i.e., water polo), and no scientific evidence illustrating the throwing efficacy (velocity and accuracy) and repeatability of PAPE activities have on the overhead throw. This study compared the effects of two different PAPE activities, seated resistance band pulldown (SRBP) and prone isometric floor push (PIFP), have on water polo throwing performance and explored the intervention’s repeatability. The study hypothesised that: i) Both PAPE activities improve maximal throwing velocity without reducing throwing accuracy; ii) they are repeatable training protocols; iii) PAPE optimal latency period is individualised. Eight national men’s water polo players were recruited to participate in the randomised crossover study design on the two PAPE intervention sessions and the repeatability sessions. Participants performed their routine warmup before completing their pre-test shots, prescribed PAPE intervention, and post-test shots for comparison. Repeated measures ANOVA was conducted to determine if maximal throwing velocity and throwing accuracy differ between both PAPE activities, however no significant differences were observed for both PAPE activities in improving throwing velocity or accuracy. Intraclass Correlation Coefficient (ICC) was conducted to evaluate repeatability of the interventions and showed good reliability and moderate-excellent validity range. Data trends showed individual improvements in throwing velocity, suggesting that PAPE responders should incorporate interventions into their warmup to improve throwing performance since intervention repeatability is reliable.
Cited by 0
Yi-Wei Lin, Malgorzata Marek-Sadowska, W. Maly et al. · 2008
In this paper, we study the circuits built from super-regular, high-density transistor arrays that can be prefabricated and customized using an OPC-free interconnect manufacturing process. The super-regular layout style greatly enhances the chippsilas manufacturability. Unlike other regular fabrics that sacrifice area and performance to improve regularity, the new layout style, combined with a new 3-D geometry transistor, enables to produce circuits with timing and power density comparable to or better than that of conventional CMOS circuits and using less chip area.
Cited by 21