Cheng-Ta Yeh

Multi-objective optimization for stochastic computer networks using NSGA-II and TOPSIS

Network reliability is a performance indicator of computer/communication networks to measure the quality level. However, it is costly to improve or maximize network reliability. This study attempts to maximize network reliability with minimal cost by finding the optimal transmission line assignment. These two conflicting objectives frustrate decision makers. In this study, a set of transmission lines is ready to be assigned to the computer network, and the computer network associated with any transmission line assignment is regarded as a stochastic computer network (SCN) because of the multistate transmission lines. Therefore, network reliability means the probability to transmit a specified amount of data successfully through the SCN. To solve this multiple objectives programming problem, this study proposes an approach integrating Non-dominated Sorting Genetic Algorithm II (NSGA-II) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). NSGA-II searches for the Pareto set where network reliability is evaluated in terms of minimal paths and Recursive Sum of Disjoint Products (RSDP). Subsequently, TOPSIS determines the best compromise solution. Several real computer networks serve to demonstrate the proposed approach.

Maximal network reliability with optimal transmission line assignment for stochastic electric power networks via genetic algorithms

This study proposes a genetic algorithm based method integrating the minimal paths and the recursive sum of disjoint products to find maximal network reliability with optimal transmission line assignment for a stochastic electric power network. In our problem, a set of transmission lines is ready to be assigned to branches of the electric power network. Because each transmission line combined with several physical lines has multiple states, the capacity of the electric power network associated with any transmission line assignment is stochastic. Network reliability is the probability that the network can transmit d units of electric power from an electric power generator (origin) to a specific area (destination). The discussed problem exhibits the features of network reliability and assignment problems, and thus it is non-deterministic polynomial-time hard. A simple electric power network and a real one are adopted to demonstrate the efficiency of the proposed algorithm while comparing with several approaches.

Maximal network reliability for a stochastic power transmission network

Many studies regarded a power transmission network as a binary-state network and constructed it with several arcs and vertices to evaluate network reliability. In practice, the power transmission network should be stochastic because each arc (transmission line) combined with several physical lines is multistate. Network reliability is the probability that the network can transmit d units of electric power from a power plant (source) to a high voltage substation at a specific area (sink). This study focuses on searching for the optimal transmission line assignment to the power transmission network such that network reliability is maximized. A genetic algorithm based method integrating the minimal paths and the Recursive Sum of Disjoint Products is developed to solve this assignment problem. A real power transmission network is adopted to demonstrate the computational efficiency of the proposed method while comparing with the random solution generation approach.

Evaluation of Optimal Network Reliability Under Components-Assignments Subject to a Transmission Budget

Network reliability evaluation for flow networks is an important issue in quality management. Many real-life systems can be modeled as stochastic-flow networks, in which each branch is multistate due to complete failure, partial failure, maintenance, etc. That is, each branch has several capacities with a probability distribution, and may fail. Hence, network reliability is the probability that a specified flow can be transmitted through the network successfully. Although there are many researches related to the evaluation of network reliability for a stochastic-flow network, how to assign a set of multistate components to the network so that the network reliability is maximal is never discussed. Therefore, this paper devotes to evaluating the optimal network reliability under components-assignments subject to a transmission budget, in which the transmission cost depends on each components capacity. The network reliability under a components-assignment can be computed in terms of minimal paths, and state-space decomposition. Subsequently, we propose an optimization method based on a genetic algorithm. The experimental results show that the proposed method can be executed efficiently in a reasonable time.

Computer network reliability optimization under double-resource assignments subject to a transmission budget

This paper discusses a double-resource assignment problem to maximize network reliability for a computer network. The resources are separated into two types: one is transmission line and another is transmission facility. In particular, each resource is multistate due to full failure, partial failure, or maintenance. Such a network assigned with multistate resources is usually modeled as a stochastic-flow network. Furthermore, each resource should have a transmission cost in reality. Hence, the network reliability is the probability that a specified demand is transmitted through the network successfully subject to a transmission budget. This paper devotes to find out the optimal double-resource assignment with maximal network reliability. An optimization algorithm combining the genetic algorithm, the minimal paths, and the Recursive Sum of Disjoint Products is developed to solve the proposed problem. The experimental results show that the proposed algorithm can be executed in a reasonable time.

Using minimal cuts to optimize network reliability for a stochastic computer network subject to assignment budget

In our modern society, information and data are usually transmitted through a computer network. Since the computer networks reliability has a great impact on the quality of data transmission, many organizations devote to evaluating or improving network reliability, especially for network reliability optimization. This study focuses on such a confronted problem that is to find the optimal transmission line assignment to the computer network such that network reliability is maximized subject to the budget constraint. Each transmission line owns several states due to failure, maintenance, etc., and thus the computer network associated with any transmission line assignment is called a stochastic computer network. Network reliability is the probability that the computer network can transmit the specified units of data successfully. Because the discussed problem is NP-hard, an optimization algorithm that integrates the genetic algorithm, minimal cuts and Recursive Sum of Disjoint Products is proposed. Experimental results illustrate the solution procedure and show that the proposed algorithm can be executed in a reasonable time.

Network reliability with deteriorating product and production capacity through a multi-state delivery network

Delivery process is a critical issue from the viewpoint of supply chain management. However, in the delivery process, deterioration would occur because of natural disasters (e.g. earthquake, fire, flood, hurricane and landslide) or man-made factors (e.g. explosion, terrorist attack and vehicular collision). The results in the intact products arriving at the market may not satisfy the demand. This paper thus concentrates on a multi-state delivery network (MSDN) with multiple suppliers, in which a vertex denotes a supplier, a transfer station or a market, while a branch denotes a carrier providing the delivery service for a pair of vertices. The available capacity of the carrier responsible for the delivery on a branch is multi-state because the capacity may be partially reserved by other customers. The addressed problem is to evaluate the network reliability, the probability that the MSDN with the deterioration consideration can satisfy the market demand within the budget and production capacity limitations. An algorithm is developed in terms of minimal paths to evaluate the network reliability along with a numerical example to illustrate the solution procedure. A real case study of the deteriorating auto glass is utilised to demonstrate the utility of the proposed algorithm.

Determine the optimal carrier selection for a logistics network based on multi-commodity reliability criterion

From the perspective of supply chain management, the selected carrier plays an important role in freight delivery. This article proposes a new criterion of multi-commodity reliability and optimises the carrier selection based on such a criterion for logistics networks with routes and nodes, over which multiple commodities are delivered. Carrier selection concerns the selection of exactly one carrier to deliver freight on each route. The capacity of each carrier has several available values associated with a probability distribution, since some of a carriers capacity may be reserved for various orders. Therefore, the logistics network, given any carrier selection, is a multi-commodity multi-state logistics network. Multi-commodity reliability is defined as a probability that the logistics network can satisfy a customers demand for various commodities, and is a performance indicator for freight delivery. To solve this problem, this study proposes an optimisation algorithm that integrates genetic algorithm, minimal paths and Recursive Sum of Disjoint Products. A practical example in which multi-sized LCD monitors are delivered from China to Germany is considered to illustrate the solution procedure.

Reliability evaluation of a stochastic-flow distribution network with delivery spoilage

From the supply chain management perspective, this paper focuses on evaluating network reliability of a stochastic-flow distribution network (SFDN) under the delivery spoilage consideration. An SFDN is composed of nodes and routes, where each node denotes a supplier, a transfer center, or a market, and each route connects a pair of nodes. Along each route, there is a carrier whose available capacity is stochastic. Moreover, goods may rot or be spoilt during delivery due to traffic accidents, collisions, natural disasters, weather, time, etc., and thus the intact goods may not satisfy the market demand. Network reliability is defined as the probability that the SFDN can satisfy the market demand under the delivery spoilage consideration and the delivery budget constraint, and can be regarded as a performance index for distribution activity in supply chain management. An algorithm is developed in terms of minimal paths to evaluate network reliability. A numerical example is given to illustrate the solution procedure. Then a practical case of fruit distribution is presented to emphasize the management implication of network reliability.

Optimal resource assignment to maximize multistate network reliability for a computer network

Computer network is a major tool to transmit data in our modern society. How to evaluate and enhance network reliability is thus an important issue for organizations, especially to maximize network reliability. A computer network is a multistate network in which each edge has several possible capacities with a probability distribution and may fail. The multistate network reliability is the probability that the maximal flow is no less than a given demand. From the standpoint of quality management, a further problem is to reassign the existing resources for maximizing multistate network reliability without changing the network topology. Hence, this paper focuses on the resource assignment problem to propose an efficient approach based on the simple genetic algorithm. In which, a resource assignment is represented as a chromosome and the corresponding multistate network reliability is the fitness value of the chromosome. The experimental results show that the proposed algorithm can derive the optimal resource assignment in a reasonable time.