Research On Decision-making Method Based On Intuitionistic Fuzzy Sets Theory For Multi-constrained Network Routing

Driven by network applications, markets, and services, the multi-constrained network routing problem based on service quality becomes a hot topic. As the network systems becomes more complex and dynamic, existing routing algorithms cannot fully support multi-constrained routing under uncertainty, ensure the quality of service (QoS) effectively, and improve resource utilization. Therefore, the intelligent decision-making method based on the uncertainty theory to solve the routing problem may open a new door to research in routing problems for new generation information network.Focusing on the transformation of multi-constrained routing problem with uncertain network state parameters into the multi-attribute decision making problem based on intuitionistic fuzzy sets (Vague sets), several projects on this topic are conducted:First, the problems of existing routing algorithms based on fuzzy logic. Second, the approach to fuzzy process the QoS metrics is provided. Thirdly, to identify the optimal path with a single constraint, a shortest path mathematical model is build and two algorithms are developed to solve the shortest path problem based on intuitionistic fuzzy sets (Vague sets). Then two algorithms are developed based on possibility degree matrix and score functions, respectively, to solve the weighted composite constrained and independent multi-constrained routing problems, respectively. A novel multi-constrained routing decision making method is also build to identify a more reasonable path. This method is built on the similarity measures among vague values and vague sets based on multiple geometric characteristics and on similarity measures of vague sets. Finally, the applications programme and experiments of the decision-making method are provided.The main contribution of the research is summarized as follows.In recognizing the existence of uncertain parameters for the network status and multiple constraints for the QoS routing problem, the flaws of existing routing algorithms are identified based on fuzzy logic (including the inaccurate information representation, the unreasonable operator selection, and the incomplete set of route selection rules) through the research on the construction and selection of fuzzy implication operators. Using intuitionistic fuzzy sets which are more reasonable than traditional fuzzy sets to represent the uncertain network state information, methods are provided to fuzzy process the QoS metrics, to name a few, hop, delay, bandwidth, and residual energy.Through the analysis of fuzzy shortest path problems, a mathematical model is developed including mathematical description and definitions of the fuzzy operations for QoS metrics and the routing decision-making. Based on this analysis, the SM-SPP and its improved algorithm SM-SPP+is presented to find the shortest path with single-constrain. The vague length of all possible paths from the source node to the destination node is calculated in the classical directed network, by mean of vague similarity measure to evaluate similarity degree between the vague path lengths and the theoretical shortest path length. The path with the highest similarity degree is the shortest path. Two illustrative examples are given to demonstrate the results of the approaches.In the study of multi-attribute decision model and multi-constrained routing, two multi-attribute decision-making methods are used based on possibility degree matrix and score function, respectively, to solve the routing problem. First, a composite additive constraint with vague set is represented, set attribute weights and generate possible matrix according to "shortest path" principle, and obtain the ordering of the paths by the ranking formula. Second, the QoS metrics and the selection criteria of the optimal path is determined according to the network environment and the traffic stream, develop the evaluation function and the score function, express each path as a vague set using seven fuzzy quantification, calculate and compare the degree of similarity of each path length and the selection criteria, the path with the highest degree of similarity is the optimal path. Two illustrative examples are given to demonstrate the correctness and feasibility of the algorithm.With the study of geometric representation of vague sets, a novel similarity measure between vague values and between vague sets is developed based on multiple geometric characteristics including area and center distance. The result of the numerical experiments shows that the proposed similarity measure is more intuitive, effective and reasonable than most of the existing similarity measures in general. According to the criteria of "The nearer to the optimal, the better, the degree of similarity measure is defined and more rational decisions can be obtained. An approach is proposed for multi-constrained routing problem based on similarity measures of vague sets. This approach determines the QoS metrics with the traffic stream and the selection criteria of the optimal path and worst path, calculates the similarity measures of the theoretically determined optimal/worst path and each feasible path, and compare the degree of similarity measure of each feasible path, the the one with the highest degree of similarity is the optimal path. Several illustrative examples are given to demonstrate the correctness and feasibility of the proposed approach. The feasibility and correctness of the approach is validated in the Matlab platform using a large amount of randomly generated data. The routing decision simulation system (RDSS) is developed using Java programming language. Through the different types of streams routing decisions and transmission in the RDSS, the experiment shows that the SM-MCR algorithm is self-adaptive in the dynamic network environment. Comparing to the Greed-Delay algorithm and energy algorithm MBCR, the proposed approach has a lower loss rate and the total energy consumption is more reasonable. When the data flow increases, the load is more balanced. The results also show that the approach has a lower total delay than MBCR algorithm and the total delay for the approach corresponds with Greed-Delay algorithm. Finally, The open shortest path first (OSPF) and the dynamic source routing (DSR) routing protocol as two representatives, the applications programme of the multi-constrained routing system with decision method are presented based on intuitionistic fuzzy sets (Vague sets).