For long-term operation of the security alarm system of a critical infrastructure facility, it is important to maintain alarm systems with seismic sensors installed in controlled areas of its perimeter. The task of forming a maintenance strategy requires reliable data on the required number of backup alarm systems for a given duration of operation and the permissible value of the duration of the process of replacing a faulty alarm system. The article shows a stochastic model of the operational reliable behavior of a fragment of the security alarm complex with the installation of five alarm systems with seismic sensors in the controlled area of the perimeter. The configuration of five alarm systems with seismic sensors installed on the controlled areas of the perimeter provides the use of the majority principle with the rule {3 out of 5} to decide on the type of intruder. At the same time, the majority principle provides an effective way to detect an inoperable alarm system with seismic sensors, which is important for the maintenance of the security alarm system. Such a model is necessary for choosing an appropriate option for the maintenance strategy of the security alarm system, which will provide a given value of its reliability for long-term operation.

In article presented theoretic generalization practical efficiency issues training and deployment of troops (armed forces) by improving the analysis system, generalization and implementation of combat experience (ASGIE) and an actual scientific problem was solved in a new way, consisting in the improvement of the scientific and methodological apparatus of its functioning. Performed analysis mathematical models for descriptions and assessment functioning system of training and use of troops (armed forces). In the same time, the results of researches in exercises and other activities of operational, combat training and daily activities of troops (armed forces) showed the absence of effective mathematical models for the functioning of ASGIE of troops (armed forces). The constituents for each of the three performance indicators have been substantiated and determined, which reproduce the corresponding reactions of the ASGIE to obtain a generalized experience. The connection of these constituents with the requirements for the mathematical model of the reaction of the ASGIE is shown. A discrete-continuous stochastic model of the ASGIE reaction has been developed to study the effectiveness of a promising ASGIE for various options for its implementation. A system of Kolmogorov-Chapman differential equations of the first order has been formed, the solution of which gives the opportunity to get the value of performance indicators ASGIE. Calculations performed for ASGI which includes the governing bodies of three military units and three OC with decision-making on the generalization of experience for the majority principle "2 with 3" show: acceptable values of the performance indicator "Experience is generalized" are achieved when the values of the probability of identifying experience and the probability of successful generalization of experience by the management body are greater than 0.9; with the values of the probability of experience identification and the probability of successful generalization of experience by the governing bodies in the range from 0.4 to 0.7 with a probability of not less than 0.7, it will be necessary to continue to accumulate generalized experience in order to make a decision on its dissemination; accordingly, a value in the range from 0.2 to 0.6 is unacceptable for organizing the process of generalizing experience for the developed ASGI.

The peculiarity of the special purpose communication information network is that its functioning is influenced by the tactical situation. In particular, it determines the intensity of the flow of packets with information about the enemy's moving objects from reconnaissance and signaling complexes to the switching node, and from it to the control point and means of destruction, depending on the probable nature of enemy action. In combat applications, the information network operates simultaneously with fast and slow packet flows. At the same time, the packet delay in the switching node should not exceed the allowable value. Excessive packet delay results in the loss of relevance of transmitted messages and, consequently, in the loss of intelligence data. The proposed method of eliminating excessive delay involves the formation of two queues of packets in the switching node. Withdrawal of service packages from two queues is carried out using a new adaptive procedure. An adaptive switch has been introduced into the structure of the switching node, designed to withdraw packets from queues to the packet service system. When selecting a packet queue, the adaptive switch compares the number of service requests from each queue. The method assumes that the number of service requests from the fast packet flow queue determines the number of packets that are in the queue. The number of requests that are formed from the queue of slow-flow packets has two components: the number of real and the number of conditional requests. Actual applications take into account the number of packets in the queue. Real requests take into account the number of packets in the queue. Conditional requests take into account the delay time, since for each missed cycle of the packet's withdrawal from the slow flow queue, the adaptive switch generates a conditional request. Therefore, the number of requests for a slow stream grows even without packets entering the queue. A comparative study of options for implementing the adaptive procedure has been carried out. The proposed method for eliminating excessive packet delay in the switching node provides an increase in the efficiency of the functioning of a special-purpose information network as a whole.