Self-organized criticality in a computer network model

We study the collective behavior of computer network nodes by using a cellular automaton model. The results show that when the load of network is constant, the throughputs and buffer contents of nodes are power-law distributed in both space and time. Also the feature of 1/f noise appears in the power spectrum of the change of the number of nodes that bear a fixed part of the system load. It can be seen as yet another example of self-organized criticality. Power-law decay in the distribution of buffer contents implies that heavy network congestion occurs with small probability. The temporal power-law distribution for throughput might be a reasonable explanation for the observed self-similarity in computer network traffic.     

Anharmonic collective excitation in a solvable model

We apply the time-dependent variational principle, the nuclear field theory, and the boson expansion method to the Lipkin model to discuss anharmonicities of collective vibrational excitations. It is shown that all of these approaches lead to the same anharmonicity to leading order in the number of particles. Comparison with the exact solution of the Lipkin model shows that these theories reproduce quite well.     

Time-dependent intranuclear cascade model

An intranuclear cascade model with explicit consideration of the time coordinate in the Monte Carlo simulation of the development of a cascade particle shower has been considered. Calculations have been performed using a diffuse nuclear boundary without any step approximation of the density distribution. Changes in the properties of the target nucleus during the cascade development have been taken into account. The results of these calculations have been compared with experiment and with the data which had been obtained by means of a time-independent cascade model. The consideration of time improved agreement between experiment and theory particularly for highenergy shower particles; however, for low-energy cascade particles (with grey and black tracks in photoemulsion) a discrepancy remains at T ? 10 GeV.     

计算机网络的长程相关特性

针对一种计算机网络模型，利用节点排队长度累计量的均方涨落函数，研究了网络节点在时间上的长程相关特性.结果表明，随着负载的增加，网络节点数据包排队长度在时间上由自由流状态的不相关或短程相关逐渐演变为临界和拥塞时的长程相关，关联范围逐渐增大，长程关联特性开始显现.在自由流状态时，节点的不相关或短程相关，并且有一致的数值为0.5的幂指数这一典型特征.而在临界状态时，节点数据包排队长度长程相关，有大于0.5的幂指数为特征.并且随网络规模的增大，节点间的群体作用逐渐显著，幂指数呈下降趋势. 关键词： 计算机网络 相变 长程相关 幂律     

网络计算机模型下海量大数据存储系统设计

对网络计算机模型下海量大数据进行安全稳定的存储,可以提高网络计算机的使用价值,增加其使用周期。但目前的海量大数据存储方法在存储过程中,无法做到对其进行灵活高效的存储,存在大数据存储分布密度较低,存储开销较大等问题。为此,以网络计算机模型体系结构为基础,提出了一种基于ARM的海量大数据存储系统设计方法。该设计方法先利用ARM芯片对网络计算机模型下海量大数据存储系统进行硬件构造,将网络海量大数据中的可利用与不可利用数据进行分类处理,采用VISA结构根据数据分类结果对大数据存储系统软件部分进行设计,依据大数据调度模型和存储相似度特征对大数据存储的时间,质量等方面进行计算,利用计算结果对大数据传输的阈值以及分布密度进行记录,最后根据循环分段的计算方式进行冗余大数据特性压缩,并对海量大数据的常规数据和冗余数据进行存储。实验仿真证明,所提方法提高了海量大数据存储的兼容性,增强了大数据存储的精确性和灵活性。     

Azimuthal correlation and collective behavior in nucleus-nucleus collisions

Physics of Atomic Nuclei - Various flow effects of nuclear and hadronic origin are investigated in nucleus-nucleus collisions. Nuclear emulsion data collected from 84Kr + Ag/Br interaction at an...     

Time-dependent density-functional theory in massively parallel computer architectures: the OCTOPUS project

Octopus is a general-purpose density-functional theory (DFT) code, with a particular emphasis on the time-dependent version of DFT (TDDFT). In this paper we present the ongoing efforts to achieve the parallelization of octopus. We focus on the real-time variant of TDDFT, where the time-dependent Kohn-Sham equations are directly propagated in time. This approach has great potential for execution in massively parallel systems such as modern supercomputers with thousands of processors and graphics processing units (GPUs). For harvesting the potential of conventional supercomputers, the main strategy is a multi-level parallelization scheme that combines the inherent scalability of real-time TDDFT with a real-space grid domain-partitioning approach. A scalable Poisson solver is critical for the efficiency of this scheme. For GPUs, we show how using blocks of Kohn-Sham states provides the required level of data parallelism and that this strategy is also applicable for code optimization on standard processors. Our results show that real-time TDDFT, as implemented in octopus, can be the method of choice for studying the excited states of large molecular systems in modern parallel architectures.     

Irregular collective behavior of heterogeneous neural networks

We investigate a network of integrate-and-fire neurons characterized by a distribution of spiking frequencies. Upon increasing the coupling strength, the model exhibits a transition from an asynchronous regime to a nontrivial collective behavior. Numerical simulations of large systems indicate that, at variance with the Kuramoto model, (i) the macroscopic dynamics stays irregular and (ii) the microscopic (single-neuron) evolution is linearly stable.     

Q-phonon scheme in the collective nuclear model

The Q-phonon scheme developed on the basis of the algebraic collective nuclear model is discussed. It is shown that, within this scheme, low-lying collective states of even-even nuclei can be described to a precision higher than 90% of the norm by using one or, at maximum, two components of the Q-phonon basis constructed by applying a fixed number of the quadrupole operators Q to the exact ground state of the system. Various applications of this approximate scheme are discussed. It is shown that, by using this scheme, the relations between several E2-transition probabilities or between the energies of the collective states can be derived. It is also shown that the Q-phonon scheme can be used to extract information about the equilibrium shapes of nuclei and their fluctuations from data on the E2-transition probabilities.     

Effect of anti-virus software on infectious nodes in computer network: A mathematical model

An e-epidemic model of malicious codes in the computer network through vertical transmission is formulated. We have observed that if the basic reproduction number is less than unity, the infected proportion of computer nodes disappear and malicious codes die out and also the malicious codes-free equilibrium is globally asymptotically stable which leads to its eradication. Effect of anti-virus software on the removal of the malicious codes from the computer network is critically analyzed. Analysis and simulation results show some managerial insights that are helpful for the practice of anti-virus in information sharing networks.     

Time-dependent Gutzwiller approximation for the Hubbard model

We develop a time-dependent Gutzwiller approximation (GA) for the Hubbard model analogous to the time-dependent Hartree-Fock (HF) method. This new formalism incorporates ground state correlations of the random phase approximation (RPA) type beyond the GA. Static quantities like ground state energy and double occupancy are in excellent agreement with exact results in one dimension up to moderate coupling and in two dimensions for all couplings. We find a substantial improvement over traditional GA and HF+RPA treatments. Dynamical correlation functions can be computed and are also substantially better than HF+RPA ones and obey well behaved sum rules.     

p t -Multiplicity correlations in a multi-pomeron-exchange model with string collective effects

The 〈p _t 〉N _ch - N _ch correlations experimentally observed in the central rapidity region in pp and p collisions, in a wide energy range from the ISR to Tevatron, are described in the framework of a multi-Pomeron exchange model in which string collectivity has been included in an effective way. Three parameters are obtained from the fit to data: the string tension, the average number of particles per string, and a parameter which effectively introduces string collective effects. The model successfully reproduces the rise of 〈p _t 〉 of charged particles, the flattening with growing rapidity density of charged particles and with the collision energy, and the negative p _t -N _ch correlation at low energies. The string tension and the average number of particles per string are energy independent, while the parameter that includes effectively string collective effects shows a smooth increasing behavior with energy. The text was submitted by the authors in English.     

虚拟网络行为对互联网整体特性的影响

虚拟网络是一种依赖Internet基础设施所提供的传输能力，但又具有独立拓扑结构和信息传递规则的应用层网络行为逻辑网络.提出了耦合虚拟网络行为与物理节点的抽象模型，对一类典型的虚拟网络逻辑拓扑给互联网整体特性带来的影响进行了分析.研究表明在虚拟网络作用下，节点数据包排队长度存在相变特性，但相变临界点比对规则网络发生了明显左移，网络性能相对恶化.当数据包注入速率小于相变临界速率时，节点数据包排队长度不相关或短程相关；在接近临界速率处，节点数据包排队长度长程相关，幂指数H增大，网络获得更强的长程相关性.同时，在注入速率大于或等于临界速率时，虚拟网络行为使网络呈现出一致的长程相关特性.     

Dynamical collective model: Even-even nuclei

A unified approach to describing the properties of spherical, transition, and deformed even-even nuclei (without invoking the concept of static nuclear deformation), a “dynamical collective model,” is set out in retrospective. We present the results of calculations for a wide range of nuclei and, on this basis, reveal a number of experiments that are of fundamental importance in developing the theory further.     

Current theoretical approaches to collective behavior of dislocations

Plastic deformation is a highly dissipative process that induces a variety of patterns such as the cell structure in multislip conditions, the matrix structure and the persistent slip bands in cyclic deformation, as also the static and propagating bands in constant strain rate conditions. The diversity and the complexity of these patterns with length scales ranging from nanometers all the way to millimeters level, and time scales ranging from picoseconds to a few hours, pose serious challenges for modeling the collective behavior of dislocations. While a large body of knowledge has accumulated on the mechanics of dislocations and their interactions for a long time, describing such patterns has been slow mainly due to lack of methods to deal with the collective behavior of dislocations. The purpose of this review is to present the rich variety of dislocation patterns observed in different deformation conditions along with the recent advances in modeling using borrowed techniques traditionally used in condensed matter physics. These can be classified as statistical and dynamical approaches. The review begins with a summary of different types of patterns and their characterization. Appropriate background material is provided both in terms of basic dislocation mechanisms and theoretical methods. The latter includes the Langevin and distribution function approaches, and a host of standard dynamical methods such as the Ginzburg–Landau approach, methods of characterization of chaos and slow manifold analysis. Statistical models for the cell structure and persistent slip bands are based on Langevin dynamics and distribution function theoretic approaches. Of the dynamical models, the first set addresses the slowly emerging matrix structure and persistent slip bands. The second set of models is devoted to the study of a type of propagative instability called the Portevin–Le Chatelier effect. Generic features of the instability addressed include bistability, negative strain rate sensitivity of the flow stress, different types of bands, the dynamics and statistics of stress drops, and their characterization. Three different models all of which are dynamical in nature are discussed. While these models are quite different with regard to their frameworks, what they seek to describe and the levels of sophistication undertaken, these models capture a good variety of the observed features. The review ends with a summary and outlook.