Search

Data center networks provide the physical infrastructure that hosts Internet‐related services and cloud computing. Designing data center networks properly is imperative for Internet‐related service and cloud computing providers to gain competitive edges through cost efficiency and service quality. In this study, we formulate a mathematical programming model to address the data center network design problem, in which the objective is to minimize total operating cost and the service delay penalty by optimizing data center location, footprint allocation, and resource provisioning decisions, while incorporating essential features, such as latency, power, multiple resources, configuration limits, and interdependent footprints. We employ a queueing model to approximate the service latency and provide tractable reformulations. To enhance computational efficiency for large‐scale problems, we further develop Lagrangian relaxation methods and generate strengthening cuts by exploiting the structural properties of the problem. Our numerical studies demonstrate that the proposed model, which jointly optimizes location, allocation, and resource provisioning, can achieve significant cost reductions and improvements in service quality compared with a hierarchical approach that optimizes these decisions sequentially. Moreover, our proposed solution methods outperform state‐of‐the‐art commercial software in terms of computational efficiency. Based on real‐world datasets, the proposed model selects data centers that have been chosen by major cloud computing infrastructure providers. We also draw managerial insights that can be used as design guidelines in practice. [ABSTRACT FROM AUTHOR]

Problem definition: In this paper, we study a reliable hub location model, the objective of which is to minimize the associated costs plus the penalty for unserved demands. The model assigns to each origin–destination pair a primary path and a backup path to hedge against the risk of random disruptions. Aside from the fixed costs of locating hubs, the fixed costs incurred by hub-connecting arcs that exhibit economies of scale in transportation are considered. Academic/practical relevance: The widely adopted hub-and-spoke architecture in the network designs can trigger cascading effects during and in the aftermath of disruption events and lead to further losses. By incorporating a general specification of disruptions, our model helps firms improve network reliability while inheriting the benefits of routing via hubs. Therefore, this work extends the literature on hub location problems by considering reliability under disruptions. Methodology: We present a path-based model and formulate a combinatorial optimization problem. By exploiting the structural properties of the problem, we introduce a tractable mixed-integer linear program reformulation and develop a constraint generation method to accelerate the solution procedure. We further construct two effective heuristics. Results: We demonstrate the necessity of considering disruptions and the profound benefits of employing backup paths. Moreover, we show via numerical studies that the path-based model delivers efficient and effective reliable hub-and-spoke designs. Managerial implications: We find that establishing backup paths according to the proposed model barely escalates network costs yet significantly enhances service levels, and that exploiting disruption correlation helps reduce the expected total cost especially when the underlying disruption correlation is strong. We also discover that the decisions of our model on the locations of hubs and hub arcs are relatively robust compared with those on path assignments. We summarize key insights as observations that can be used directly as rules of thumb to guide designs in practice. [ABSTRACT FROM AUTHOR]

This study introduces a challenging service network design problem with stochastic demands and fixed routes. In this problem, the routes of some service vehicles are fixed during the whole planning horizon, while the remaining vehicles are flexible in that their routes can adapt to different realizations of customer demands each day. The problem is to determine which vehicles should be set as fixed and how their fixed routes should be designed. To solve this problem, we formulate a two-stage stochastic mixed-integer linear program. The fixed routes are designed in the first stage before customer demands are realized, and the routes for flexible vehicles are designed in the second stage after customer demands are observed. A learning-based multiple scenario approach is developed as the solution method. Numerical experiments on real-world operational data from a logistics company show that fixing the routes of some service vehicles may reduce the operational cost by an average of approximately 6.37% compared with that in the case of no fixed routes. • The service network design problem with stochastic demands and fixed routes is introduced. • A two-stage stochastic mixed-integer linear programming model is formulated. • A learning-based multiple scenario approach is developed. • The model and solution algorithm are demonstrated on real-world operational data. [ABSTRACT FROM AUTHOR]

We introduce and study a challenging service network design problem arising within a city logistics company based in Beijing, China. To provide cost-effective same-day delivery services, the company divides the whole urban area into regions. Within each region, a fleet of small-capacity vehicles picks up and delivers packages. A fleet of line-haul vehicles (large-capacity vehicles) executes consistent multiple trips over the regions to transport consolidated packages, during which transshipment is allowed. In this study, we focus on designing the trips of line-haul vehicles, i.e., their region visiting sequences, as the performance of the logistics system depends significantly on the coordination of line-haul vehicles. We model the trip design problem as a service network design problem with consistent multiple trips. We develop a continuous-time MIP model to formulate the problem and optimally solve small-size instances. In addition, we propose an adaptive large neighborhood search heuristic to solve larger instances. Numerical experiments on randomly generated instances as well as real-world operational data demonstrate the effectiveness and efficiency of our solution approach. • The service network design problem with consistent multiple trips is introduced. • A continuous-time mixed integer programming model is formulated. • An adaptive large neighborhood search solution algorithm is developed. • The model and solution algorithm are demonstrated on real-world operational data. [ABSTRACT FROM AUTHOR]