On reliability of linear wireless sensor networks

Many solutions in the field of architecture of the Internet of things are based on the results of wireless sensor networks (WSN) research. Currently, corporations and government agencies, especially in the US, Europe and the Middle East, are making serious efforts to research, develop and patent wireless sensor network-based technologies for monitoring long pipelines for various purposes. WSNs are able to carry out continuous monitoring of critical infrastructure objects, detect any anomalies in real time and deliver the necessary data to the decision-making center. However, in this case, it becomes necessary to solve a number of problems, such as rational placement of nodes, optimization of energy consumption, efficient organization of data flows, etc., for the solution of which it is necessary to use methods for assessing the reliability and lifetime of the WSN. After all, these networks can be operated in harsh climatic conditions, network nodes can be located in places that are difficult to access for regular inspection. The main role of sensor nodes in the WSN used for monitoring is to periodically collect and transmit data to an intelligent central base station (sink, sink), where the collected data is processed to detect various anomalous events. In large- scale WSNs used to monitor long objects, such as pipelines, most sensor nodes are geographically distant from the base station and are usually equipped with self-contained low-cost batteries, the capacity of which is relatively small. This circumstance largely determines the main disadvantages of extended WSNs, since the periodic transmission of raw data over long distances through several hops to the base station leads to a rapid discharge of the battery of sensor nodes and reduces the life of the network. Other disadvantages include low reliability of wireless channels, high cost of bandwidth, low level of data security. Thus, there is a need to solve optimization problems for extended WSNs, where the objective function or constraints contain a reliability indicator. Authors propose an approach to solving a number of problems that arise when monitoring long pipelines using wireless sensor networks. A typical scenario used in a number of recent publications on this topic is considered. The WSN consists of several sensor nodes located on the surface of the pipe, the topology is a simple chain, at the end of which there is a base station. The sensors are responsible for collecting data, periodically sending packets to the base station. All nodes play an important role in data forwarding, the node closest to the base station transmits data directly to the sink, intermediate nodes are used to transmit packets from other nodes, i.e. data sent by the sender to the sink is relayed by nodes located between the sender and the sink. The main power consumption is caused by traffic transmission. Obviously, this leads to excessive waste of energy of the sensor nodes located closer to the base station due to the high asymmetric load on these nodes. Each sensor node performs periodic monitoring within its sensitivity range. All sensor nodes are initially in the same conditions, have similar communication capabilities, power consumption, their behavior is described by the same conceptual models. Each node transmits its packet to the neighboring node in the direction of the base station. The corresponding software module deployed on the base station receives data from all sensor nodes, decides on the presence or absence of a problem on its own, or via a highly reliable IP network, the data is transmitted to the decision center. To formalize the reliability criteria, the corresponding mathematical models based on Markov processes have been developed. The obtained results make it possible to find a compromise between the reliability of the monitoring network and the costs of its deployment and maintenance.