The research subject of the article is modern intellectualization methods for wireless sensor networks (WSN). Typically, their application field is uncertain and poorly formalized environments, therefore, it is possible to achieve the desired efficiency of such complexes mainly by improving the intellectual component of their control system as a whole and a separate node in particular. However, it should be noted that there is a gap between primitive behavioral models of artificial entities, for example, in swarm robotics, their interaction models and expectations from practice. The situation is aggravated by the requirements of secrecy, miniaturization, and low power consumption. In practice, it is required that a network node to be an autonomous node with the property of intelligent behavior, which also has to be able to learn the situation and make decisions, both independently taking into account the data received from other network devices and as part of a group.

The paper identifies the main requirements for the autonomous operation of an intelligent WSN. The highlighted key factors are low energy consumption and the exchange of reliable information about the environmental state to form right decisions. The purpose of the paper is to form a theoretical and mathematical base from the existing methods of WSN intellectualization that meet the above requirements, as well as to formulate proposals for further research in order to be applied in practice.

The development of new algorithms, methods and technologies that help reduce energy consumption in Internet of Things networks is a relevant task. Sensor nodes of the Internet of things consume limited energy resources when performing computing operations, receiving and transmitting data. One of the well-known approaches to reduce the energy consumption of sensor nodes is network clustering. The head node of the cluster assumes the functions of a relay of data from sensor nodes.

The paper proposes an algorithm that develops the idea of clustering networks of the Internet of things. The algorithm is based on bee swarm intelligence that assumes determining the current round cluster head and potential heads of clusters for subsequent rounds of the cycle immediately at the be-ginning of the cycle. Thus, the phase of choosing the cluster head node becomes redundant starting from the second round of the cycle, and sensor nodes do not need to perform energy-intensive calcula-tions associated with choosing the cluster head.

The simulation results show the superiority of the bee swarm algorithm in comparison with the well-known LEACH adaptive clustering algorithm with low energy consumption in terms of the duration of a wireless sensor network.