Discrete Capacity Assignment in IP networks using Particle Swarm Optimization

This paper presents a design methodology for IP networks under end-to-end Quality-of-Service (QoS) constraints. Particularly, we consider a more realistic problem formulation in which the link capacities of a general-topology packet network are discrete variables. This Discrete Capacity Assignment (DCA) problem can be classified as a constrained combinatorial optimization problem. A refined TCP/IP traffic modeling technique is also considered in order to estimate performance metrics for networks loaded by realistic traffic patterns. We propose a discrete variable Particle Swarm Optimization (PSO) procedure to find solutions for the problem. A simple approach called Bottleneck Link Heuristic (BLH) is also proposed to obtain admissible solutions in a fast way. The PSO performance, compared to that one of an exhaustive search (ES) procedure, suggests that the PSO algorithm provides a quite efficient approach to obtain (near) optimal solutions with small computational effort.     

Large deviations of Markovian polling models with applications to admission control

In this paper we consider large deviations and admission control problems for a discrete-time Markovian polling system. The system consists of two-parallel queues and multiple heterogeneous servers. The arrival process of each queue is a superposition of mutually independent Markovian on/off processes, and the multiple servers serve independently the two queues according to the so called Bernoulli service schedule. Using the large deviations techniques, we derive upper and lower bounds of the overflow probabilities, and then we present an admission control criterion by which different Quality of Service (QoS) requirements for the two queues are guaranteed.     

State-Dependent Coupling in General Networks

In this paper we extend a result which holds for the class of networks of quasireversible nodes to a class of networks constructed by coupling Markov chains. We begin with a network in which the transition rates governing the stochastic behaviour of the individual nodes depend only on the state of the node. Assuming that the network has an invariant measure, we construct another network with transition rates at each node depending on the state of the entire network, and obtain its invariant measure.     

Performance of a Leaky Bucket System with Long-Range Dependent Input Traffic

The leaky bucket is a flow control mechanism that is designed to reduce the effect of the inevitable variability in the input stream into a node of a communication network. In this paper we study what happens when an input stream with heavy tailed work sessions arrives to a server protected by such a leaky bucket. Heavy tailed sessions produce long-range dependence in the input stream. Previous studies of the systems suggested that such long-range dependence can have dramatic effect on the system performance. By concentrating on the time until overflow of large finite buffers, we characterize the distribution of the time until buffer overflow, and we show that the leaky bucket flow control does make the system overflow less often, but long-range dependence still makes its presence felt.     

Large deviations analysis of the generalized processor sharing policy

In this paper we consider a stochastic server (modeling a multiclass communication switch) fed by a set of parallel buffers. The dynamics of the system evolve in discrete-time and the generalized processor sharing (GPS) scheduling policy of [25] is implemented. The arrival process in each buffer is an arbitrary, and possibly autocorrelated, stochastic process. We obtain a large deviations asymptotic for the buffer overflow probability at each buffer. In the standard large deviations methodology, we provide a lower and a matching (up to first degree in the exponent) upper bound on the buffer overflow probabilities. We view the problem of finding a most likely sample path that leads to an overflow as an optimal control problem. Using ideas from convex optimization we analytically solve the control problem to obtain both the asymptotic exponent of the overflow probability and a characterization of most likely modes of overflow. These results have important implications for traffic management of high-speed networks. They extend the deterministic, worst-case analysis of [25] to the case where a detailed statistical model of the input traffic is available and can be used as a basis for an admission control mechanism. This revised version was published online in June 2006 with corrections to the Cover Date.     

Active sound control in composite regions

In active sound control, noise shielding of a target region is achieved via additional sources (called controls) situated at the perimeter of the region. The sources protect the target region by adjusting the acoustic field near the boundary of the region. In the present paper a numerical model of active sound control based on surface potentials in 3D bounded composite regions is numerically studied. In the composite region setup, it is required that the regions be shielded from noise while allowing admissible sound that is generated in the shielded regions to be preserved. The admissible sound is usually required to propagate freely inside the protected regions or in a (selective) predetermined pattern. The adjusting approach used here does not require any knowledge of the sound sources or the properties of the propagation medium in order to obtain the controls. Moreover, the approach differs sharply from some other approaches where the detailed knowledge of the sound sources and the propagation medium is required. For the first time, numerical test cases involving both free communication and predetermined communication pattern between the regions in three dimensions are considered. In all test cases, these regions are effectively shielded from the noise while any present admissible sound is preserved. In addition, selective propagation of the admissible sound between the regions is enforced. The effect of the number of controls on their operation is also studied. Whether admissible sound is present or not, the level of noise cancellation decreases linearly as fewer controls are used. In addition to the increase in size of the interference zone, the controls become individually distinguishable.     

Capacity planning of a perinatal network with generalised loss network model with overflow

Recent literature shows that the arrival and discharge processes in hospital intensive care units do not satisfy the Markovian property, that is, interarrival times and length of stay tend to have a long tail. In this paper we develop a generalised loss network framework for capacity planning of a perinatal network in the UK. Decomposing the network by hospitals, each unit is analysed with a GI/G/c/0 overflow loss network model. A two-moment approximation is performed to obtain the steady state solution of the GI/G/c/0 loss systems, and expressions for rejection probability and overflow probability have been derived. Using the model framework, the number of required cots can be estimated based on the rejection probability at each level of care of the neonatal units in a network. The generalisation ensures that the model can be applied to any perinatal network for renewal arrival and discharge processes.     

Asymptotic analysis of a large closed queueing network with discriminatory processor sharing

In this paper the steady-state behavior of a closed queueing network with multiple classes and large populations is investigated. One of the two nodes of the network simply introduces random delays and the discipline in the other node is discriminatory processor sharing. The network is not product-form, so not even the steady-state behavior is known. We assume that the usage is moderately heavy, and obtain two-term asymptotic approximations to the mean number of jobs, and the mean sojourn time, of each class of jobs in the processor node. We also obtain the leading term in the asymptotic approximation to the joint distribution of the number of jobs in the processor node, which is a zero-mean multivariate Gaussian distribution around a line through the origin.     

Heuristics for Distribution Network Design in Telecommunication

A distribution network problem arises in a lower level of an hierarchical modeling approach for telecommunication network planning. This paper describes a model and proposes a lagrangian heuristic for designing a distribution network. Our model is a complex extension of a capacitated single commodity network design problem. We are given a network containing a set of sources with maximum available supply, a set of sinks with required demands, and a set of transshipment points. We need to install adequate capacities on the arcs to route the required flow to each sink, that may be an intermediate or a terminal node of an arborescence. Capacity can only be installed in discrete levels, i.e., cables are available only in certain standard capacities. Economies of scale induce the use of a unique higher capacity cable instead of an equivalent set of lower capacity cables to cover the flow requirements of any link. A path from a source to a terminal node requires a lower flow in the measure that we are closer to the terminal node, since many nodes in the path may be intermediate sinks. On the other hand, the reduction of cable capacity levels across any path is inhibited by splicing costs. The objective is to minimize the total cost of the network, given by the sum of the arc capacity (cables) costs plus the splicing costs along the nodes. In addition to the limited supply and the node demand requirements, the model incorporates constraints on the number of cables installed on each edge and the maximum number of splices at each node. The model is a NP-hard combinatorial optimization problem because it is an extension of the Steiner problem in graphs. Moreover, the discrete levels of cable capacity and the need to consider splicing costs increase the complexity of the problem. We include some computational results of the lagrangian heuristics that works well in the practice of computer aided distribution network design.     

Network flow interdiction on planar graphs

The network flow interdiction problem asks to reduce the value of a maximum flow in a given network as much as possible by removing arcs and vertices of the network constrained to a fixed budget. Although the network flow interdiction problem is strongly NP-complete on general networks, pseudo-polynomial algorithms were found for planar networks with a single source and a single sink and without the possibility to remove vertices. In this work, we introduce pseudo-polynomial algorithms that overcome various restrictions of previous methods. In particular, we propose a planarity-preserving transformation that enables incorporation of vertex removals and vertex capacities in pseudo-polynomial interdiction algorithms for planar graphs. Additionally, a new approach is introduced that allows us to determine in pseudo-polynomial time the minimum interdiction budget needed to remove arcs and vertices of a given network such that the demands of the sink node cannot be completely satisfied anymore. The algorithm works on planar networks with multiple sources and sinks satisfying that the sum of the supplies at the sources equals the sum of the demands at the sinks. A simple extension of the proposed method allows us to broaden its applicability to solve network flow interdiction problems on planar networks with a single source and sink having no restrictions on the demand and supply. The proposed method can therefore solve a wider class of flow interdiction problems in pseudo-polynomial time than previous pseudo-polynomial algorithms and is the first pseudo-polynomial algorithm that can solve non-trivial planar flow interdiction problems with multiple sources and sinks. Furthermore, we show that the k-densest subgraph problem on planar graphs can be reduced to a network flow interdiction problem on a planar graph with multiple sources and sinks and polynomially bounded input numbers.     

Hamiltonian 2-forms in Kähler geometry, III extremal metrics and stability

This paper concerns the existence and explicit construction of extremal Kähler metrics on total spaces of projective bundles, which have been studied in many places. We present a unified approach, motivated by the theory of Hamiltonian 2-forms (as introduced and studied in previous papers in the series) but this paper is largely independent of that theory. We obtain a characterization, on a large family of projective bundles, of the ‘admissible’ Kähler classes (i.e., those compatible with the bundle structure in a way we make precise) which contain an extremal Kähler metric. In many cases every Kähler class is admissible. In particular, our results complete the classification of extremal Kähler metrics on geometrically ruled surfaces, answering several long-standing questions. We also find that our characterization agrees with a notion of K-stability for admissible Kähler classes. Our examples and nonexistence results therefore provide a fertile testing ground for the rapidly developing theory of stability for projective varieties, and we discuss some of the ramifications. In particular we obtain examples of projective varieties which are destabilized by a non-algebraic degeneration.     

Sensible decisions based on QoS

Network Quality of Service (QoS) criteria of interest include conventional metrics such as throughput, delay, loss, and jitter, as well as new QoS criteria based on power utilization, reliability and security. Variable and adaptive routing have again become of interest in networking because of the increasing importance of mobile ad-hoc networks. In this paper we develop a probability model of adaptive routing algorithms which use the expected QoS to select paths in the network. Our objective is not to analyze QoS, but rather to design randomized routing policies which can improve QoS. We define QoS metrics as non-negative random variables associated with network paths which satisfy a sub-additivity condition along each path. We define the QoS of a path, under some routing policy, as the expected value of a non-decreasing measurable function of the QoS metric. We discuss sensitive and insensitive QoS metrics, the latter being dependent on the routing policy which is used. We describe routing policies simply as probabilistic choices among all possible paths from some source to some given destination. Incremental routing policies are defined as those which can be derived from independent decisions taken at certain points (or nodes) along paths. Sensible routing policies are then introduced: they take decisions based simply on the QoS of each available path. Sensible policies, which make decisions based on the QoS of the paths, are introduced. We prove that the routing probability of a sensible policy can always be uniquely obtained. A hierarchy of m-sensible probabilistic routing policies is then introduced. A 0-sensible policy is simply a random choice of routes with equal probability, while a 1-sensible policy selects a path with a probability which is inversely proportional to the (expected) QoS of the path. We prove that an m + 1-sensible policy provides better QoS on the average than an m-sensible policy, if the QoS metric is insensitive. We also show that under certain conditions, the same result also holds for sensitive QoS metrics.Accepted: May 2003, This work was supported by the U.S. Army and Navy under contracts N611339-00-K-0002, N61339-02-C0050, N61339-02-C0080, N61339-02-C0117, and by NSF grants EIA0086251, EIA0203446, ECS0216381.     

Metric techniques for convex stationary ergodic Hamiltonians

We adapt the metric approach to the study of stationary ergodic Hamilton?CJacobi equations, for which a notion of admissible random (sub)solution is defined. For any level of the Hamiltonian greater than or equal to a distinguished critical value, we define an intrinsic random semidistance and prove that an asymptotic norm does exist. Taking as source region a suitable class of closed random sets, we show that the Lax formula provides admissible subsolutions. This enables us to relate the degeneracies of the critical stable norm to the existence/nonexistence of exact or approximate critical admissible solutions.     

A parametric algorithm for convex cost network flow and related problems

The convex cost network flow problem is to determine the minimum cost flow in a network when cost of flow over each arc is given by a piecewise linear convex function. In this paper, we develop a parametric algorithm for the convex cost network flow problem. We define the concept of optimum basis structure for the convex cost network flow problem. The optimum basis structure is then used to parametrize v, the flow to be transsshipped from source to sink. The resulting algorithm successively augments the flow on the shortest paths from source to sink which are implicitly enumerated by the algorithm. The algorithm is shown to be polynomially bounded. Computational results are presented to demonstrate the efficiency of the algorithm in solving large size problems. We also show how this algorithm can be used to (i) obtain the project cost curve of a CPM network with convex time-cost tradeoff functions; (ii) determine maximum flow in a network with concave gain functions; (iii) determine optimum capacity expansion of a network having convex arc capacity expansion costs.     

Stability and performance analysis of rate-based feedback flow controlled ATM networks

Motivated by ABR class of service in ATM networks, we study a continuous time queueing system with a feedback control of the arrival rate of some of the sources. The feedback about the queue length or the total workload is provided at regular intervals (variations on it, especially the traffic management specification TM 4.0, are also considered). The propagation delays can be nonnegligible. For a general class of feedback algorithms, we obtain the stability of the system in the presence of one or more bottleneck nodes in the virtual circuit. Our system is general enough that it can be useful to study feedback control in other network protocols. We also obtain rates of convergence to the stationary distributions and finiteness of moments. For the single botterneck case, we provide algorithms to compute the stationary distributions and the moments of the sojourn times in different sets of states. We also show analytically (by showing continuity of stationary distributions and moments) that for small propagation delays, we can provide feedback algorithms which have higher mean throughput, lower probability of overflow and lower delay jitter than any open loop policy. Finally these results are supplemented by some computational results. This revised version was published online in June 2006 with corrections to the Cover Date.     

One-phase inverse Stefan problems with unknown nonlinear sources

We consider quasilinear models of inverse problems with phase transitions in a domain whose external boundary is a phase front with an unknown dependence on time. Additional information for finding the sources is given in the form of final overdetermination for the solution of the direct Stefan problem. On the basis of the duality principle, we obtain sufficient conditions for the uniqueness of the solution in Hölder classes for the considered inverse Stefan problems. We present examples in which the uniqueness property is lost when extending the set of admissible solutions.     

移位交换网的最优路由算法

移位交换网是重要的互联网络之一 ,在并行计算中有着广泛应用 .然而 ,它缺少任意点对间的最短路由算法 .已有的路由算法都不能保证其任意节点对间都是最短路由 .文中给出了一个最短路由算法 ,也是最优路由算法 ,它使得从源节点到目的节点的任何信息都是沿最短路由传输 .同时 ,我们还得到了任意节点对间的距离公式     

QoS-Guided Bandwidth Management in Differentiated Time Scales

This paper presents an integrated approach to QoS-guided network bandwidth allocation where each traffic flow requires a sufficient bandwidth allocation to support its mean traffic rate and to meet a delay requirement. Under the assumption that the peak rate of each traffic flow is decreasing with the time interval within which the rate is measured, we derive an analytical relationship between the delay bound and the bandwidth requirement for each individual flow. Then, based on a Gaussian aggregate traffic model, we show that two key controllable parameters, the coefficient of variation and the provision for variation for the aggregate traffic flow, determine all three fundamental QoS attributes (throughput, delay, and loss). We illustrate by examples that these results can be used to design admission policies. We demonstrate also quantitatively a remarkable QoS advantage of larger channel bandwidth in a statistical multiplexing environment. The analytical contributions are expected to be generally useful in QoS-guided bandwidth management in broadband networks.     

Likelihood Inference for Large Scale Stochastic Blockmodels With Covariates Based on a Divide-and-Conquer Parallelizable Algorithm With Communication

We consider a stochastic blockmodel equipped with node covariate information, that is, helpful in analyzing social network data. The key objective is to obtain maximum likelihood estimates of the model parameters. For this task, we devise a fast, scalable Monte Carlo EM type algorithm based on case-control approximation of the log-likelihood coupled with a subsampling approach. A key feature of the proposed algorithm is its parallelizability, by processing portions of the data on several cores, while leveraging communication of key statistics across the cores during each iteration of the algorithm. The performance of the algorithm is evaluated on synthetic datasets and compared with competing methods for blockmodel parameter estimation. We also illustrate the model on data from a Facebook derived social network enhanced with node covariate information. Supplemental materials for this article are available online.     

A traffic control algorithm for Clos network type ATM switches

This paper presents a new traffic control algorithm for asynchronous transfer mode (ATM) switches composed of Clos network type switch fabric. In most traffic control algorithms previously proposed, an ATM switch has been considered as a single node, although the switch fabric of an ATM switch is usually of a network type. In this paper, we suggest a new control algorithm that is designed not only for the ATM network but also for the switch fabric and that can maintain high speed even in cases where buffer capacity of the switch fabric is limited. Main modules of the suggested algorithm are a traffic status detection mechanism based on non-parametric statistical tests, a cell-loss detection mechanism, and a cluster-based fair share computation procedure. Results of simulation experiments show that the suggested algorithm satisfactorily adjusts traffic rate of available bit rate services according to changes in traffic rate used by quality of service (QoS) guaranteed services. The results also show that the algorithm prevents cell losses relatively well and keeps the delay time of QoS guaranteed services short enough.