APPLYING A FORMALIZED LOGICAL-PROBABILISTIC METHOD FOR DETERMINING THE SURVIVABILITY INDICATORS OF MESH NETWORKS

Abstract

Using the example of an LORA MESH network, the capabilities of the logic-probabilistic trajectory modeling method for determining the survivability indicators of mission-critical systems with a mesh network structure are presented. The proposed method is formalized, and based on it, software can be implemented that allows the designer to obtain quantitative values of survivability indicators at the system engineering design stage, when the system has not yet been built in hardware, through modeling. The analysis of known methods for calculating pairwise connectivity indices shows that the use of certain organized intermediate data structures can enhance the computational efficiency of algorithms. In the correlation method, such structures are the sets of paths for the i-th iteration. In the decomposition method, relative to the key element with indexed masking of elements in the set of simple paths, these structures include an indexed array of network elements, which determines their ordering by the chosen criterion, a vector of masking levels for elements in the set of simple paths, and a stack of probabilities for incomplete paths. Therefore, the question arises about the possibility of constructing and applying other auxiliary structures and investigating the efficiency of corresponding algorithms. During the conducted research, the idea emerged to use a "spanning tree" of simple paths as such an auxiliary structure, by means of which it is possible to manage the passage of "probability packets". Methods for evaluating survivability with a grid structure of communication networks are characterized by high computational complexity. Such computational complexity hinders the multivariate analysis of design decisions. The formalization of the probabilistic packet switching method has allowed reducing computational complexity for the user (enabled the implementation of this method in software, as most calculations are performed repeatedly automatically; moreover, when changing the network configuration, the function for calculating connection pairs is formed with less time expenditure). This will allow for multivariate analysis of design decisions, improve the efficiency of system design, and significantly reduce time costs.