Research On Optimization Of Wireless Sensor Networks Based On Fractal Theory

Wireless sensor networks is not a new technology, but with the promotion and popularization of the Internet of Things, as the key technology of IOT and Ubiquitous Network, sensor network has a rapid growth trend, and the network deployment spread from a single industrial environment to different areas, for example agriculture, forestry, environmental protection, health care and transportation areas. This situation makes the size and the types of data transmission in wireless sensor networks have a lot of changes. The network endures pressure coming from uneven allocation of resources and large amount of data transmitted. Particularly, it has been an unprecedented challenge in order to meet the requirements of a higher demand for real-time data processing capability and transmission capacity of wireless sensor networks. Since the wireless sensor network is considered to be a network that be capable of detecting the data, the functionality of wireless sensor network is focused on the perception of the data acquisition and transmission. The problems faced by wireless sensor networks can be divided into two types. The first is the rationality of large-scale network structure. The second is the data transmission efficiency perceived by the node data type and the large amount of data. Optimization of network structures and data transmission has become the main point of research in wireless sensor networks. Currently, the unbalance of large-scale sensor network architecture and the inefficient of data transfer is being studied. The increased size of network, in essence, is to increase network complexity. The increased complexity is not only in topology, but also in network data. When adding nodes, the links between adjacent nodes become intricately connected. To join and withdraw from any node will have an impact on the local network topology, and the node with the reduction of energy needs it to participate with data transmission in the most economical way, which reflects the complexity of the network structure. The smart sensor nodes have the ability to request the network simultaneously transmit a number of different types of data. The different data structures, the different transmission requirements, different data size, data from a single to multiple allows simultaneous transmission of different data transmission increases the complexity of the data. Thus, the research in the view of sensor network structure and the heterogeneous data, and the study of optimization of network complexity are theoretical and practical.Fractal dimension as a description of irregular space and a measurement of filling has been used for analysis of Internet and other complex network. The fractal dimension is creatively used in this paper in the analysis for the complexity of wireless sensor networks. The paper promoted the network structure dimension and the dimension of the data flow model. It also analyzes the impact of the relationship between the parameters of dimension and network performance. In addition, based on the dimension analysis, the proposed network control algorithm optimizes the network structure and data transmission. The optimized results will reduce the energy consumption and extend the node network lifecycle. This paper provides new ideas and methods for the wireless sensor network research. The main innovation of the paper is as follows:1) Verify that the characteristics of fractal of wireless sensor network architecture. The proposed fractal model (Probability Fractal Model, PFM model) is based on a simple fractal wireless sensor. The dimension structure can be calculated by using this model and the result can be used to adjust the network structure.The innovation verified the fractal characteristics of the architecture of wireless sensor network. It also verified the fractal of wireless sensor network has the characteristics of self-similar. It promoted that it is reasonable to use of the fractal dimension when research on the wireless sensor network research. On the basis of fractal characteristics of wireless sensor networks, it proposed a simple model based on fractal wireless sensor network transmission probabilities. The Markov chain model with the probability of successful transmission node in the transmission process was calculated after a successful transmission network link to get an accurate fill level in the entire transmission path, and further calculated the fractal dimension of an effective network transmission path. The calculation of fractal dimension of the wireless sensor network architecture breaks the limitation that the fractal theory which originally be used only in curve, or three-dimensional images of objects. It creatively uses the fractal theory for wireless sensor networks. Calculation of the dimension of the wireless sensor network provides a method for quantify the complexity of wireless sensor network, and an evaluation for the optimization of the network. By changing the parameters of the network nodes and the network structure, the complexity of network can be adjusted in order to optimize the network.2) Verify the multiple fractal characteristics of the wireless sensor network data stream. The paper proposed the interval sequences accumulate data flow model (Time Interval Sequence Accumulative Flow, TAF model). The model is obtained by calculating the number of different types of data streams in the complexity of network transmission. It provides the basis for traffic prediction and path selection.The innovation verified the characteristics of self-similar for data streams of wireless sensor network It verified that it is reasonable to use the multiple fractal in data stream analysis for wireless sensor network. According to the characteristics of accumulated data stream for hybrid sensor network, it proposed time-series interval of accumulated data flow model. The model can accumulate many types of data streams within the network. It calculates the accumulated data flow with the fractal dimension, and identifies multiracial function of the mixed data flow. It used the multi-fractal method for small time scales to calculate the Dimension of the wireless sensor network data stream. The method is both to meet the short-range correlation of the data stream, and consistent with the coexistence of multiple types of data characteristics. The plurality of types of real-time data stream type calculation can better predict the fluctuation in the data flow network to help predict the impact of the flow path and the key nodes. It facilitates the early preprocessing of balanced data transmission, and optimizes the transmission path.3) To optimize the energy consumption of wireless sensor networks, this paper proposes a fractal-based low-power wireless sensor networks control algorithm (Low-energy Consumption Clustering algorithm, LCC algorithm). The algorithm can effectively reduce the energy consumption of the network nodes to extend network life cycle.The control algorithms for fractal-based low-power wireless sensor networks in the actual operation included the network structure optimization and data transmission optimization in every process. Each of the optimization takes place in two different time cycles. Optimization of the network structure takes place when the clusters set up. When the structure needs to be optimized, the cluster must be reorganized according to an optimization algorithm for a better structure. Data transfer optimization may happen multiple times within a cluster round. The algorithm is based on the calculation of the transmission of data dimension in order to choose the best path for the data transmission. LCC algorithms ensure the equalizing structure and efficient transmission as it considered network structure having a long-term correlation with the data stream having a short-term correlation. The paper compares the LCC, LEACH and improved LEACH algorithm LEACH-C. The simulation results show that under the same running conditions and time, LLC algorithm ensures more nodes running. It proves that the LCC algorithm reduce the network energy consumption and optimize the network.