Server Location Research Articles

Disjunctive Decomposition for Two-Stage Stochastic Mixed-Binary Programs with Random Recourse

This paper introduces disjunctive decomposition for two-stage mixed 0-1 stochastic integer programs (SIPs) with random recourse. Disjunctive decomposition allows for cutting planes based on disjunctive programming to be generated for each scenario subproblem under a temporal decomposition setting of the SIP problem. A new class of valid inequalities for mixed 0-1 SIP with random recourse is presented. In particular, we derive valid inequalities that allow for scenario subproblems for SIP with random recourse but deterministic technology matrix and right-hand side vector to share cut coefficients. The valid inequalities are used to derive a disjunctive decomposition method whose derivation has been motivated by real-life stochastic server location problems with random recourse, which find many applications in operations research. Computational results with large-scale instances to demonstrate the potential of the method are reported.

P2P-Based Approach to Finding Replica Server Locations for Alleviating Flash Crowds

Many services rely on the Internet to provide their customers with immediate access to information. To provide a stable service to a large number of customers, a service provider needs to monitor demand fluctuations and adjust the number and the location of replica servers around the world. Unfortunately, flash crowds make it quite difficult to determine good number and locations of replica servers because they must be repositioned very quickly to respond to rapidly changing demands. We are developing ExaPeer, an infrastructure for dynamically repositioning replica servers on the Internet on the basis of demand fluctuations. In this paper we introduce ExaPeer Server Reposition (EPSR), a mechanism that quickly finds appropriate number and locations of replica servers. EPSR is designed to be lightweight and responsive to flash crowds. EPSR enables us to position replica servers so that no server becomes overloaded. Even though no dedicated server collects global information such as the distribution of clients or the load of all servers over the Internet, the peer-to-peer approach enables EPSR to find number and locations of replica servers quickly enough to respond to flash crowds. Simulation results demonstrate that EPSR locates high-demand areas, estimates their scale correctly and determines appropriate number and locations of replica servers even if the demand for a service increases/decreases rapidly.

Optimal Content Distribution in Video-on-Demand Tree Networks

Video-on-demand (VOD) services are expected to grow significantly over time, providing diverse programs for home entertainment, distance instruction, news on demand, and other applications. These services require large bandwidth resources. We present a model for a tree network where each node may have demands for multiple different VOD program families, where each program family is an aggregation of similar programs. The decision variables include location of VOD servers and assignment of program families to servers. The model considers cost of servers, cost of assigning program families to servers, and cost of link bandwidths used to broadcast programs. The objective is to minimize the sum of all these costs. The model is formulated as an integer program. We develop a dynamic programming formulation with multiple state variables and an algorithm that solves this model. Starting from the end nodes of the tree network, the algorithm determines optimal solutions to subtrees, eventually reaching the root node with the central server, thus providing an optimal solution to the entire network.

Algorithms for Server Placement in Multiple-Description-Based Media Streaming

Multiple description coding (MDC) has emerged as a powerful technique for reliable real-time communications over lossy packet networks. In its basic form, it involves encoding a media stream into r substreams that are sent independently from a source to a destination. Each substream (or description) can be decoded independent of the other r-1 substreams. With every successful reception of a substream, the quality of the decoded signal improves. In this paper, we consider the problem of placing a set of servers in the network such that a desired quality of service can be provided to a community of clients that request MDC-coded traffic. We formulate the server placement (SP) problem, with the goal of identifying the minimum number of server locations that can provide r descriptions to a set of clients such that the delay associated with each path from a chosen server location to a given client is bounded by a given delay constraint and the total ldquounreliabilityrdquo associated with the group of paths to a given client is also upper bounded. We show that the SP problem belongs to the class of NP-complete problems. We propose a mixed-integer linear programming (MILP) formulation and an efficient heuristic solution for the SP problem. Simulations are conducted to evaluate the performance of the proposed algorithm and compare it with the optimal solution provided by the MILP solution.

Locating repair servers in hierarchical reliable multicast networks to reduce the makespan

AbstractIn a hierarchical reliable multicast (HRM) environment, makespan is the time that is required to fully and successfully transmit a packet from the sender to all receivers. Low makespan is vital for achieving high throughput with a TCP‐like window‐based sending scheme. In HRM methods, the number of repair servers and their locations influence the makespan. In this paper we propose a new method to decide the locations of repair servers that can reduce the makespan in HRM networks. Our method has a formulation based on mixed integer programming to analyze the makespan minimization problem. A notable aspect of the formulation is that heterogeneous links and packet losses are taken into account in the formulation. Three different heuristics are presented to find the locations of repair servers in reasonable time in the formulation. Through simulations, three heuristics are carefully analyzed and compared on networks with different sizes. The results show that the our best heuristic is close to the lower bound by a factor of 2.3 in terms of makespan and by a factor of 5.5 in terms of the number of repair servers. Copyright © 2008 John Wiley & Sons, Ltd.

On the optimal allocation of users to servers in an emergency service

This paper considers the optimal allocation of users to servers in an emergency medical service, in which two or three ambulances cooperate to serve calls from a region of rectangular shape. Queueing models are often used to approach emergency health services. We analyse three different dispatch policies to determine the best available unit to be sent and introduce a classification of districts in four different classes. Properties of the optimal boundaries, i.e., the ones that minimize the average response time between users and servers and which define districts, given the home locations of servers, are investigated. A two-server model is reviewed, and results are obtained for a three-server model, enlightening a non-monotone behaviour of the shift of the optimal boundaries, with respect to the equidistance curve, in the different states of the system, depending on the utilization ratio, on the policy assumed and on the locations of the servers. This way, it is not possible to give an a priori' general rule on the properties of the boundaries of the districts only on the basis of the number of busy servers at the specific time.

Online chasing problems for regular polygons

We consider a server location problem with only one server to move. In this paper we assume that a request is given as a region and that the service can be done anywhere inside the region. Namely, for each request an online algorithm chooses an arbitrary point in the region and moves the server there. Note that if every request is a single point and the server must exactly go there in the given order as conventional server problems, there is no choice for the online player and the problem is trivial. Our main result shows that if the region is a regular n-gon, the competitive ratio of the greedy algorithm is 1 / sin π 2 n for odd n, and 1 / sin π n for even n. In particular for a square region, the greedy algorithm turns out to be optimal.

Server placement and path selection for QoS‐enabled overlay networks

AbstractMultimedia services are in strong need of end‐to‐end Quality of Service (QoS) and multicasting support. This need cannot be fulfilled by the current Internet yet, due to difficulties associated with implementing these techniques ubiquitously across the numerous autonomous systems of this global network. Overlay Networks offer a means to implement these advanced network features in a logical overlay layer on top of the Internet. The problem of finding the locations of the overlay servers, coupled with the paths that are selected, has not yet been studied in sufficient detail. Therefore, we present a formal multilayer formulation of the server location and path selection problem for these overlay networks as an Integer Linear Programming (ILP) problem. An algorithm based on standard ILP solution techniques and a number of heuristics were developed. These algorithms are compared with each other and with standard Internet routing in terms of acceptance rate, end‐to‐end delay and usage of the network bandwidth. Copyright © 2008 John Wiley & Sons, Ltd.

A comparative study of decomposition algorithms for stochastic combinatorial optimization

This paper presents comparative computational results using three decomposition algorithms on a battery of instances drawn from two different applications. In order to preserve the commonalities among the algorithms in our experiments, we have designed a testbed which is used to study instances arising in server location under uncertainty and strategic supply chain planning under uncertainty. Insights related to alternative implementation issues leading to more efficient implementations, benchmarks for serial processing, and scalability of the methods are also presented. The computational experience demonstrates the promising potential of the disjunctive decomposition (D 2) approach towards solving several large-scale problem instances from the two application areas. Furthermore, the study shows that convergence of the D 2 methods for stochastic combinatorial optimization (SCO) is in fact attainable since the methods scale well with the number of scenarios.

A multiple server location–allocation model for service system design

Service systems are endemic in a service economy, and effective system design is fundamental to the competitiveness of service organizations such as retailers, distributors, and healthcare providers. This is because system design may significantly facilitate (or hinder) the attainment of important organizational objectives such as minimizing system cost and maximizing service level. This paper develops and solves a comprehensive nonlinear location–allocation model for service system design that incorporates several relevant costs and considerations. These include, for instance, transportation, facility, and waiting costs, queuing considerations, multiple servers, multiple order priority levels, multiple service sites, and service distance limits. The model is first converted to an equivalent linear form and then solved using Lagrangian relaxation. A computational study shows problems with 250 service districts, 60 service sites, and 250 candidate locations are solved in about two and a half minutes. An extensive managerial experiment is conducted that evaluates alternative system designs from a number of important perspectives including centralization versus decentralization, system configuration, and service distance limit. Each scenario is evaluated with respect to two fundamental criteria, namely, total cost and service level. The analysis provides insights into important tradeoffs that must be taken into consideration in designing an effective service system.

Improving the resilience of content distribution networks to large scale distributed denial of service attacks

Distributed Denial of Service (DDoS) attacks remain a daunting challenge for Internet service providers. Previous work on countering these attacks has focused primarily on attacks at a single server location and the associated network infrastructure. Increasingly, however, high-volume sites are served via content distribution networks (CDNs). In this paper, we propose two mechanisms to withstand and deter DDoS attacks on CDN-hosted Web sites and the CDN infrastructure. First, we present a novel CDN request routing algorithm which allows CDN proxies to effectively distinguish attacks from the requests from actual users. The proposed scheme, based on the keyed hash function, can significantly improve the resilience of CDNs to DDoS attacks. In particular, the resilience of a CDN, consisting of n proxies, becomes O(n^2) with the proposed approach, when compared to a site hosted by a single server. We present performance numbers from a controlled test environment to show that the proposed approach is effective. Second, we introduce novel site allocation algorithms based on the well-established theory on binary codes. The proposed allocation algorithm guarantees an upper bound on the level of service outage of a CDN-hosted site even when a DoS attack on another site on the same CDN has been successful. Together, our schemes significantly improve the resilience of the Web sites hosted by CDNs, and complement other work on countering DoS.

The multiple server location problem

In this paper, we introduce the Multiple Server location problem. A given number of servers are to be located at nodes of a network. Demand for these servers is generated at each node, and a subset of nodes need to be selected for locating one or more servers in each. There is no limit on the number of servers that can be established at each node. Each customer at a node selects the closest server (with demand divided equally when the closest distance is measured to more than one node). The objective is to minimize the sum of the travel time and the average time spent at the server, for all customers. The problem is formulated and analysed. Results using heuristic solution procedures: descent, simulated annealing, tabu search and a genetic algorithm are reported. The problem turns out to be a very difficult combinatorial problem when the total demand is very close to the total capacity of the servers.

Focus on software

Focus on software

ATM network design for corporate networks

Asynchronous transfer mode (ATM) networks play an important role in support of distributed applications and modern corporate systems. The applications that use these architectures transfer high volumes of data in high-speed bursts and have stringent delay requirements. To address these needs, network managers seek cost efficient network solutions that provide network capacity sufficient to support high-speed applications. In this paper we present a new formulation and solution procedure for designing ATM networks to support corporate applications. Given the locations of the application servers and multiple clients we would like to design a minimum cost ATM network by which the clients can access the servers. Most of the previous work done in this area separates the problem into a routing problem (virtual circuit routing) and an assignment (virtual path assignment) problem. Thus, the solutions tend to be sub-optimal in the combined problem. This research optimizes virtual path (VP) assignment and virtual circuit (VC) routing simultaneously. We formulate the combined problem explicitly and develop an effective solution method. The solution method provides the designer with virtual paths and virtual circuits over which the actual communication takes place. Computational results are provided.

A fuzzy queuing location model with a genetic algorithm for congested systems

This article presents a fuzzy location–allocation model for congested systems. In service networks, such as health and emergency services, public safety, fire fighting and so on, the location of servers and allocation of demand nodes to them have a strong impact on the congestion at each server and as such, on the quality of service. The previous efforts in this area have concentrated on enhancing the reliability and quality of service with a probabilistic orientation. In this paper we utilize fuzzy theory to develop a queuing maximal covering location–allocation model which we call the fuzzy queuing maximal covering location–allocation model. We consider fuzzified queuing parameters as well as fuzzified constraints to develop a new mathematical model which we convert to a single objective integer programming model. Our model considers one type of service call, one type of server and includes one constraint on the quality of service in the form of a service time or a queue length constraint. A genetic algorithm is developed to solve and test the model using up to 50-node networks. We also propose extensions to our model.

The Million-Variable “March” for Stochastic Combinatorial Optimization

Combinatorial optimization problems have applications in a variety of sciences and engineering. In the presence of data uncertainty, these problems lead to stochastic combinatorial optimization problems which result in very large scale combinatorial optimization problems. In this paper, we report on the solution of some of the largest stochastic combinatorial optimization problems consisting of over a million binary variables. While the methodology is quite general, the specific application with which we conduct our experiments arises in stochastic server location problems. The main observation is that stochastic combinatorial optimization problems are comprised of loosely coupled subsystems. By taking advantage of the loosely coupled structure, we show that decomposition-coordination methods provide highly effective algorithms, and surpass the scalability of even the most efficiently implemented backtracking search algorithms.

네트워크 소요시간 최소화를 위한 이동 에이전트의 멀티캐스트 이주 모델 구현

이동 에이전트는 호스트 간의 통신 횟수, 전송 데이터의 양, 에이전트의 크기, 네트워크 상태 등의 요소에 따라 매우 다양한 성능을 보이며 특히, 이주 방식은 분산 시스템의 전체 성능에 큰 영향을 준다. 대부분의 기존 이주 방식은 고정된 순서대로 호스트를 방문하고 방문한 호스트에서 작업을 수행한 후, 결과를 계속 누적시키면서 이주하는 단순 구조를 가지고 있다. 그러므로 호스트의 결점 및 장애, 서비스 부재 등과 같은 상황이나 방문해야할 호스트의 수가 많을 경우, 네트워크 소요시간이 증가되어 이동 에이전트를 사용하는 것 자체가 비효율적일 수 있다. 따라서 본 논문에서는 이러한 문제들을 해결하여 네트워크 소요시간을 최소화하기 위한 멀티캐스트 이주 모델을 설계 및 구현한다. 멀티캐스트 이주 모델은 분산된 서버의 위치 투명성 및 에이전트 객체 복제 정보를 제공하는 네이밍 에이전트와 호출 모듈만을 포함한 이동 에이전트 등과 같은 요소를 포함한다. 그리고 구현된 이주 모델을 검증하기 위해 프로토타입 시스템에 적용하여 기존 이주 방식과 비교 평가한다. The mobile agent has very various performance according to the element of communication number of times between hosts, quantity of transmission data agent's size, network state etc. specially, migration method is caused much effect in whole performance of distributed system. Most existing migration methods have simplicity structure that it moves doing to accumulate continuously result after achieving task by visiting host in the fixed order. Therefore, in case there are situation such as fault, obstacle, and service absence etc. This can be inefficient due to mobile agent increased network required time. In this paper, we design and implementation Multicast Migration Model for minimizing network required time by solving this problems. Multicast Migration Model includes components such as mobile agent including call module and naming agent, which provides object replication information and distributed server's location transparence. And we evaluate and compare with existing migration method applying prototype system to verify implemented migration model.

Process for improving GPS acquisition assistance data and server-side location determination for cellular networks

This paper introduces a process for analyzing and improving the quality of the acquisition assistance data produced for Assisted GPS (A-GPS) positioning in cellular networks. An experimental test bed is introduced and a series of experiments and results are provided. The experiments validate the GPS acquisition assistance data in open-sky conditions. Accuracy results for initial testing of our server location determination engine in a range of different environments are also given with results of a long-term run. Acquisition assistance data provides the GPS handset with information that allows it to detect the GPS signals more quickly and allows detection of much weaker signals. It does this by providing information to the handset about where to look for the signals. The AGPS server is a mobile location server determining the location of devices within a cellular network. In order to measure the quality of the acquisition assistance data produced by the A-GPS server a piece of hardware has been developed called the “A-GPS Trainer”. This takes assistance data and uses it to lock on to the satellites. It then provides code phase measurements back to the server for it to do a location calculation. The trainer also reports on the amount of time it takes to lock on to the signals of the individual satellites. It can run multiple calculations over a period of time and report the results.

Performance of adaptive CORBA middleware

Middleware provides inter-operability and transparent location of servers in a heterogeneous distributed environment. A careful design of the middleware software is required, however, for achieving high performance. This research proposes an adaptive middleware architecture for CORBA-based systems. The adaptive middleware agent that maps an object name to the object reference has two modes of operations. In the handle-driven mode, it returns a reference for the requested object to the client that uses this reference to re-send the request for the desired operation to the server, whereas in the forwarding mode it forwards the entire client request to the server. The server upon invocation performs the desired operation and returns the results to the client. An adaptive ORB dynamically switches between these two modes depending on the current system load. Using a commercial middleware product called Orbix-MT, we have implemented a skeletal performance prototype for the adaptive ORB. Based on measurements made on a network of workstations and a synthetic workload we observe that the adaptive ORB can produce a significant benefit in performance in comparison to a pure handle-driven or a pure forwarding ORB. Our measurements provide valuable insights into system behavior and performance.

Towards unified formulations and extensions of two classical probabilistic location models

We give a unified view of Daskin's Maximum Expected Covering Location Problem (MEXCLP) and ReVelle and Hogan's Maximum Availability Location Problem (MALP), identifying similarities and dissimilarities between these models and showing how they relate to each other. These models arise in the location of servers in congested emergency systems. An existing extension of MEXCLP is reviewed; we then develop an extension of MALP and give the corresponding mathematical formulation. These two extensions are obtained when the simplifying assumptions of the original models are dropped and Larson's hypercube model is embedded into local search methods. In this paper these methods are further enhanced by the use of simulated annealing. Computational results are given for problems available in the literature.