Maintainable Stochastic Flow Networks with High QoS: A Quick and Practical Approach

摘要

In practically all network flow problems, attaining the desired quality of service (QoS) while maintaining minimum possible transmission time is one of the most important challenges. Since most of these networks are stochastic in nature, ensuring QoS becomes a major concern, especially when the network maintenance cost is also required to be kept low. Lin [25] has recently proposed a system reliability evaluation method for a maintainable stochastic flow network (MSFN). The problem is to find out the most reliable set of disjoint paths from a given source to a sink, such that d units of data can be communicated in T units of time, within a specified maintenance budget B. Data is distributed among disjoint paths for parallel communication. Although, the method provides the most reliable set of disjoint paths through which data can be communicated within the time and maintenance constraints, it takes enormous time in determining the most reliable set of disjoint paths within a given network. This leads to significant system overhead in deciding the most reliable paths and results in tighter time bound. Keeping this in view, a fast solution to data communication problem is proposed in the present paper that ensures high system reliability for communicating d units of data within T units of time under a given maintenance budget. The method is compared with that proposed in [27] and is shown to be much faster without any compromise in the quality of service (QoS).