Connectivity Modeling And Optimization Of Energy Harvesting Wireless Sensor Networks

With the development and maturity of manufacturing technology, the application of wireless sensor networks has become more and more widely. But the traditional sensor nodes are usually powered by batteries, which greatly limit the lifetime of the network, therefore, energy harvesting technology is an effective way to solve the problem of energy supply in wireless sensor networks. In order to improve the performance of wireless sensor networks, etwork connectivity is a hot issue to be studied.In energy harvesting wireless sensor networks, since directly depending on environmental factors, the amount of sensor nodes' collected energy is a stochastic process in the period of time, and this randomness of energy amount leads to the randomness of whether the sensor nodes can work. Furthermore, for energy saving considerations, sensor nodes generally use Duty-Cycle scheduling strategy, a link can be established only when the two nodes are active at the same time, thus in EH-WSNs, whether the link between nodes can be established is probabilistic. In view of above characteristics and the constraint of Energy harvesting wireless sensor networks on communication radius, In this paper, we establish the EW-random graph model by using the idea of ER-random graph, and then, the most common sink node centric network in practical application is analyzed based on this model, and according to the random graph model of EW, the network connectivity model based on probability is established. In this model, we take the minimum probability of node connectivity, the average value and the maximum value as well as the network's maximum connectivity ratio as a measure of the network connectivity, and three methods are proposed to calculate the probability of node connectivity, the first one is sub graph probabilistic method based on random complete graph,the second is path probability method based on random complete graph, and the last is simulation statistics method based on probability. Based on SEW random graph model, the paper establishes the relationship between network connectivity and node density, node probability, link probability, communication radius and sink node position by means of simulation experiments, and the optimization strategy of network connectivity is put forward according to each factor.At the end of the paper, an energy harvesting wireless sensor network simulation system is implemented, the prototype system consists of energy harvesting wireless sensor network simulation platform and the data collection system based on the e Z430-RF2500-SHE development tool. The simulation platform has realized the deployment of energy harvesting wireless sensor networks, connectivity computing and simulation experiments. And the data collection subsystem implements a voltage detection and data acquisition system based on the real energy harvesting sensor nodes. And based on the analysis of voltage data, the probability of energy collection is verified.