Sensor network connectivity models
Keywords:
wireless sensor network, probability-time characteristics, probability-energy characteristics, connectivity probability, message delivery time, cluster, signal power at the transmitting antenna, network connectivity, μ-quantile of delivery tiAbstract
Introduction: One of the key areas in the research of wireless sensor networks is studying the ways to increase the battery life
by saving energy in individual devices. The article introduces and discusses a new energy-efficient stochastic measure of the quality
parameter for a wireless sensor network – connectivity, which reflects the ability of a network to establish connections between
its elements within the boundaries of the sensor field in real time, at a certain level of the sensor device battery charge. Purpose:
Identifying the interdependence between the probability-time and probability-energy characteristics, as well as the influence, on these
characteristics, of such parameters as geometric dimensions, distribution model of sensor devices within the sensor field, network
topology and message routing algorithms. Results: A new stochastic characteristic of wireless sensor network functioning quality is
proposed, called connectivity. It encompasses the spatial, temporal and energy characteristics of the network, making it possible to
describe, from a general point of view, a wide range of problems which arise when you study the functioning of wireless networks
at the stages of data collecting, distributing and processing by the sensors. Stochastic connectivity indicators are introduced for
wireless sensor networks, describing a network as a whole and allowing you to investigate the delay and blocking of the information
exchange, taking into account the size of the sensor field and the power consumed by individual devices. Models are built for assessing
the probability of wireless sensor network connectivity, message delivery time and delivery time quantile, improving the accuracy
of network quality assessments. Practical relevance: The obtained models and methods are supposed to be used in digitalization of
agricultural organizations and in the educational process at Knyagininsky University.