Международный журнал информационных и коммуникационных технологий
НАУЧНО-ПРАКТИЧЕСКАЯ КОНФЕРЕНЦИЯ «30-ЛЕТИЕ НЕЗАВИСИМОСТИ КАЗАХСТАНА
жүктеу/скачать 3,05 Mb. Pdf көрінісі
|
| НАУЧНО-ПРАКТИЧЕСКАЯ КОНФЕРЕНЦИЯ «30-ЛЕТИЕ НЕЗАВИСИМОСТИ КАЗАХСТАНА:
ДОСТИЖЕНИЯ И ПЕРСПЕКТИВЫ» Material and research methods Wireless sensor networks are self-organizing networks consisting of many miniature sensor nodes that monitor phenomena, processes, environmental characteristics, etc. [23]. The WSN sensor nodes probe a given coverage area, collecting information, process it, and then transmit the processed data to the receiving node. All nodes are monitored and controlled by the base station. The plane, on which the WSN is located, is called a sensor field. The number of sensor nodes in such a field can reach several tens of thousands of units; therefore, the traditional methods of organizing infrastructure networks are not suitable for the implementation of the WSN. Self-organization of the network allows the use of sensor nodes as needed, for example, when a node has detected an event. The rest of the time, the sensory nodes are in a dormant state, which makes it possible to maintain their energetic capabilities, which are very limited. Therefore, they are designed to consume as little power as possible, since they often operate in an unknown environment and replacement of the power supply may not be possible. The task of saving energy in WSNs is one of the most important, since a lower energy consumption for transmitting and receiving information allows extending the functioning of a sensor network, that is, extending the WSN life cycle. To increase this duration, it is better to use a cluster organization, because the number of sensory nodes in one sensory field can be very large. Clustering saves energy, since data transmission is limited between several nodes, which increases the life of a sensor network [24]. Formation of clusters in a dynamic environment is one of the main problems of hierarchical routing, as it affects the energy consumption in the WSN. When using clustering, head cluster nodes (HCNs) are selected in each cluster, they are used to collect data from the cluster nodes and transfer them to the base station. A node with any level of residual energy, including the minimum, can be selected as the HCNs. However, a large number of headend nodes with low energy can lead to the network failure. That is why the characteristics of HCNs determine the performance of the cluster and the entire WSN, and the HCNs choice plays an essential role in functioning of sensor networks. The choice of head units and routing in the sensor network is carried out using algorithms. The basic algorithm for the WSN is LEACH (Low Energy Adaptive Cluster Hierarchy). It provides a probabilistic selection of the sensor node for the role of the head at the beginning of the functioning of the sensor network and rotation based on the energy characteristics of the sensor nodes. This extends the life cycle of sensor nodes and the network as a whole, but does not provide optimal coverage for a long time. For example, clustering of the BSS based on the basic LEACH algorithm has made it possible to increase the life cycle of the sensor network seven times compared to a conventional sensor field [25]. It should be noted that this algorithm performs only direct data transfer within the cluster and directly from the head node of the WSN, which is not always possible due to the large size of the network. Many algorithms are mainly aimed at increasing the duration of the functioning of sensor nodes and the network as a whole. For example, the HEED algorithm uses a hybrid criterion for head node selection based on residual energy analysis and the location of nearby nodes. The PEGASIS algorithm forms a coherent structure by connecting the farthest from the base station nodes to the closest to the base station nodes, whereas the search for clusters is an iterative process during which the estimated center of the cluster is calculated and displaced until it coincides with the center of a mass of a local cluster of objects. As a result, the entire set of original objects is divided into subsets based on the proximity of elements to each other. For each subset of the cluster, its center is determined, where the macro cloud is located. Then, it will be minimally removed from all cluster members. The placement of clusters with this approach has been obtained using the FOREL clustering algorithm. The essence of this method is to search for traffic consumer’s clusters, the number of which is close to a given value of the cluster radius. Figure 1 shows the architecture of WSNs, where head nodes can be both stationary and mobile devices, such as UAVs. Clusters are formed within the range of the head units, overlapping the coverage areas of radio communication. FSNs can also consist of mobile sensor nodes, which significantly complicates the problem of choosing a head node and cluster stability during the operation of the network. In accordance with typical models, the speed of movement of such nodes does not exceed 2 m/s. And even at this speed, it can go out of range of the head node of the cluster before the end of collecting and transferring data to this head node. International Journal of Information and Communication Technologies, Special issue, march 2022 53 жүктеу/скачать 3,05 Mb. Достарыңызбен бөлісу:
|