A survey on resource allocation in high performance distributed computing systems

An efficient resource allocation is a fundamental requirement in high performance computing (HPC) systems. Many projects are dedicated to large-scale distributed computing systems that have designed and developed resource allocation mechanisms with a variety of architectures and services. In our study, through analysis, a comprehensive survey for describing resource allocation in various HPCs is reported. The aim of the work is to aggregate under a joint framework, the existing solutions for HPC to provide a thorough analysis and characteristics of the resource management and allocation strategies. Resource allocation mechanisms and strategies play a vital role towards the performance improvement of all the HPCs classifications. Therefore, a comprehensive discussion of widely used resource allocation strategies deployed in HPC environment is required, which is one of the motivations of this survey. Moreover, we have classified the HPC systems into three broad categories, namely: (a) cluster, (b) grid, and (c) cloud systems and define the characteristics of each class by extracting sets of common attributes. All of the aforementioned systems are cataloged into pure software and hybrid/hardware solutions. The system classification is used to identify approaches followed by the implementation of existing resource allocation strategies that are widely presented in the literature.     

Transient effects on microchannel electrokinetic filtering with an ion-permselective membrane

The electrokinetics and hydrodynamics in a hybrid microfluidic/nanofluidic pore network configuration and its effect on the concentration enrichment of charged analytes are described. A hydrogel microplug, photopolymerized in a microfluidic channel, with negative surface charge serves as a nanoporous membrane and dictates the electrokinetic behavior within the adjoining microchannel compartments. The nanoporous hydrogel with a mean pore size on the order of the electrical double layer thickness imparts ion-permselectivity (cation-selectivity) to the migration of ionic species which, under the influence of an applied electrical field, drives concentration polarization in bulk solution near the interfaces between the two microchannel compartments and the hydrogel-based nanopores. The concentration enrichment efficiency for charged analytes depends on this concentration polarization, which strongly affects the distribution of local electrical field strength. In addition, electroosmotic flow in the device plays a critical role in determining the location of the analyte enrichment zone. A theoretical model and simulations are presented to explain the interplay of concentration polarization and electroosmotic flow with respect to the observed concentration enrichment of negatively charged analytes at the cathodic hydrogel plug-microchannel solution interface.     

DAWL: A Delayed-ACK Scheme for MAC-Level Performance Enhancement of Wireless LANs

The IEEE 802.11 MAC protocol provides a reliable link layer using Stop & Wait ARQ. The cost for high reliability is the overhead due to acknowledgement packets in the direction opposite to the actual data flow. In this paper, the design of a new protocol as an enhancement of IEEE 802.11 is proposed, with the aim of reducing supplementary traffic overhead and increasing the bandwidth available for actual data transmission. The performance of the proposed protocol is evaluated through comparison with IEEE 802.11 as well as with a SSCOP-based protocol. Results underline significant advantages of the proposed protocol against existing ones, thus confirming the value and potentiality of the approach.Dzmitry Kliazovich received his Masters degree in telecommunication science from Belarusian State University of Informatics and Radioelectronics in 2002. He is currently working towards the Ph.D. degree in University of Trento, Italy. His main research interest lies in wireless networking field with a focus on performance optimization and cross-layer design.Fabrizio Granelli was born in Genoa in 1972. He received the “Laurea” (M.Sc.) degree in Electronic Engineering from the University of Genoa, Italy, in 1997, with a thesis on video coding, awarded with the TELECOM Italy prize, and the Ph.D. degree in Electronic Engineering and Computer Science from the same university in 2001. Since 2000 he is carrying on his teaching activity as Assistant Professor at the Dept. of Information and Communication Technologies (DIT) of the University of Trento (Italy) within the B.Sc. and M.Sc. Degrees in Telecommunications Engineering.The research interests of Dr. Granelli are mainly focused on networking, with particular attention to network modeling and performance evaluation, wireless networks, access control, and next-generation telecommunication networks.He is author of more than 30 refereed papers, published in several international journals and conferences.Dr. Granelli is member of the IEEE Committee on “Communication Systems Integration and Modeling” (CSIM) and of the Technical Programme Committee of the “QoS and Performance Evaluation Symposium” of the International Conference on Communications (ICC 2003 and ICC 2004).     

NUMERICAL SIMULATION OF ELECTROKINETIC MICROFLUIDICS IN COLLOIDAL SYSTEMS

A numerical scheme is presented for simulating electrokinetic microfluidics in systems with arbitrary morphology. This scheme is based on a numerical solution of the coupled Poisson, Nernst–Planck, and Navier–Stokes equations. While traditional finite-difference methods were used to resolve the first two problems, the lattice Boltzmann method was applied to the latter. The developed numerical approach was used for the simulation of electroosmotic flow through a simple cubic array of hard (impermeable, nonconducting) micro-sized spheres. Volumetric electroosmotic flow was studied for dependence on electrical field strength, ζ-potential at the solid-liquid interface, electrical double layer interaction, and numerical grid resolution. Colloid stability and electrokinetics in microfluidic devices with particulate or monolithic fixed-bed elements represent two potential applications of this work.     

豌豆淀粉/ε-聚赖氨酸复合膜的制备及其性能

本文以豌豆淀粉/ε-聚赖氨酸为主要基材,以甘油、海藻酸钠等为增塑剂制备复合膜包装材料。采用单因素、正交实验等方法,在设定的工艺条件下流延成膜,进行复合膜的配方研究和最佳条件下膜性能研究。结果表明:最佳配方为100 mL水中加入10.5 g豌豆淀粉、1.5 g聚赖氨酸(调pH9)、1.80%甘油、0.50%海藻酸钠;产品性能检测表明,抗拉强度约为18.35 MPa,断裂延伸率约为34.79%,水蒸气透过率为12.10 g/(m2·h),透油率为0.682 g·m/(m2·h),对大肠杆菌具有较好的抑制性。本研究表明豌豆淀粉/ε-聚赖氨酸复合膜具有功能性包装应用的基础。     

Propagating concentration polarization and ionic current rectification in a nanochannel-nanofunnel device

We study ionic current rectification observed in a nanofluidic device with a nanofunnel positioned between two straight nanochannels. Ion transport is simulated by resolving the coupled three-dimensional Nernst-Planck, Poisson, and Navier-Stokes equations. In the modeled system, the electric double layer extends into the channel, and consequently, the funnel tip exhibits charge-selective properties, which results in the formation of enriched and depleted concentration polarization (CP) zones within the nanofunnel in the high- and low-conductance states, respectively. This scenario is similar to the one observed for ion transport through a charged conical nanopore connecting two macroscopic reservoirs. However, the presence of the adjacent straight nanochannels allows the CP zones to propagate out of the funnel into the adjoining channels. The condition for propagation of the CP zones is determined by several parameters, including the electroosmotic flow velocity. We demonstrate that in the high-conductance regime the modeled system is characterized by increased ionic concentrations in the entire cathodic nanochannel, whereas in the low-conductance state the depleted CP zone does not propagate out of the funnel and remains localized. The required three-dimensional modeling scheme is implemented on a parallel computational platform, is general as well as numerically efficient, and will be useful in the study of more advanced nanofluidic device designs for tailoring ionic current rectification.     

A metamodeling approach for cross-layer optimization: A framework and application to Voice over WiFi

Cross-layer design has been proposed to optimize the performance of networks by exploiting the interrelations among parameters and procedures at different levels of the protocol stack. This paper introduces a quantitative framework for the study of cross-layer interactions, which enables design engineers to analyze and quantify interlayer dependencies and to identify the optimal operating point of the system, by using network economic theory principles. The framework is then used for performance optimization of a single-cell Voice over WiFi (VoWiFi) system. Insights gained from the considered scenario enable us to define a novel cross-layer Call Admission Control (CAC) scheme. The multistage CAC takes into account Quality of Service (QoS) criteria, which provide satisfaction to the end user, as well as revenue criteria that maximize the possible profit of the WiFi provider.     

Pore-scale dispersion in electrokinetic flow through a random sphere packing

The three-dimensional velocity field and corresponding hydrodynamic dispersion in electrokinetic flow through a random bulk packing of impermeable, nonconducting spheres are studied by quantitative numerical analysis. First, a fixed bed with interparticle porosity of 0.38 is generated using a parallel collective-rearrangement algorithm. Then, the interparticle velocity field is calculated using the lattice-Boltzmann (LB) method, and a random-walk particle-tracking method is finally employed to model advection-diffusion of an inert tracer in the LB velocity field. We demonstrate that the pore-scale velocity profile for electroosmotic flow (EOF) is nonuniform even under most ideal conditions, including a negligible thickness of the electrical double layer compared to the mean pore size, a uniform distribution of the electrokinetic potential at the solid-liquid interface, and the absence of applied pressure gradients. This EOF dynamics is caused by a nonuniform distribution of the local electrical field strength in the sphere packing and engenders significant hydrodynamic dispersion compared to pluglike EOF through a single straight channel. Both transient and asymptotic dispersion behaviors are analyzed for EOF in the context of packing microstructure and are compared to pressure-driven flow in dependence of the average velocity through the bed. A better hydrodynamic performance of EOF originates in a still much smaller amplitude of velocity fluctuations on a mesoscopic scale (covering several particle diameters), as well as on the microscopic scale of an individual pore.     

CA-DAG: Modeling Communication-Aware Applications for Scheduling in Cloud Computing

This paper addresses performance issues of resource allocation in cloud computing. We review requirements of different cloud applications and identify the need of considering communication processes explicitly and equally to the computing tasks. Following this observation, we propose a new communication-aware model of cloud computing applications, called CA-DAG. This model is based on Directed Acyclic Graphs that in addition to computing vertices include separate vertices to represent communications. Such a representation allows making separate resource allocation decisions: assigning processors to handle computing jobs, and network resources for information transmissions. The proposed CA-DAG model creates space for optimization of a number of existing solutions to resource allocation and for developing novel scheduling schemes of improved efficiency.