Research On Traffic Analysis And Traffic Control Technologies In Wireless Sensor Networks

As the rise and development of the Internet of things technology, the new applications and the number of nodes and users are going to emerge to grow explosively. The internet of things gateway and core network are coming into face the impact of the mass data under the trend that the applications of wireless sensor network come into large scale. It is predictive and very significant to focus on the fundamental research of traffic analysis and traffic control of the wireless sensor networks in order to optimize the network resources and enhance the network performance under the mass data. We have analyzed the traffic characteristics and studied the effect of the self-similar traffic from the perspectives of traffic analysis, and then raised resource scheduling solutions from the perspectives of traffic control. The main innovation ideas as follows:1) According to the trend of the wireless sensor networks traffic characteristics and the network traffic characteristics analysis methods of internet and communications network, the distribution characteristics of the inter-arrival time are fitted and analyzed, and then the burst characteristic is verified in the multi-scale time dimensionality; The self-similar characteristics of the inter-arrival time and network traffic are verified by V-T method and R/S analysis; It is revealed that the root cause of the wireless sensor networks self-similarity is heavy-tailed distributions of the inter-arrival time based on the summary and analysis of the reasons of the network self-similar characteristics currently.2) The accumulative combination model based on IAT-H modeling inter-arrival time in wireless sensor networks with cycle work mode is proposed according to the influence of traffic characteristics, whose simulation data have similar mean, variance and the self-similar characteristics compared with the real data; Then ATBH method based on accumulative combination IAT-H model is proposed for traffic synthesis, the network traffic with different work mode and data type are obtained used to analyze the network self-similarity; Finally the influence of the gateway performance caused by wireless sensor networks traffic is researched under the traffic input obtained using ATBH method, we find that the gateway self-similarity is more obvious than the input traffic with the big buffer and energy conservation mode.3) In order to solve the self-similar traffic problem by allocating the resources inside a node, a traffic control method based on cross-layer resource optimization is researched:The cross-layer algorithm of resource optimization based on network utility and network lifetime is proposed; The energy control in physical layer, access control in MAC layer and traffic control in transfer layer are considered together to represent the optimization problem, and the vertical decomposing is used by introducing congestion price and energy price to update the energy, link access probability and rate; The series of resolution is distributed, and simulation results prove. the convergence of the resolution which achieve the stability of the network traffic.4) In order to solve the self-similar traffic problem by allocating the resources among the nodes of the networks, a traffic control method based on dynamic multi-path routing is researched:A queuing model is firstly established particularly under the condition of finite buffers and dynamic routing for WSN; A routing algorithm (LSR) which is designed to throughput stability for WSN by using Lyapunov methodology is proposed; Based on this, a clustering distributed routing strategy which can balance the load of the nodes in their cluster is proposed by using LSR way; Finally, the simulation results show that this series of routing mechanism can effectively implement the load balance and the traffic stability of WSN.