| Various disasters and accidents occur frequently in China,which are presented in different types,cause wide-ranging impacts and bring out heavy economic losses,etc.The rapid construction of temporary communication networks to collect emergency data is of great significance for preventing,reducing and relieving disasters.Wireless sensor networks are often used for data collection in sensing scenarios,but the limited sensing range due to fixed node location and the high latency disadvantage of multi-hop back transmission method make it difficult to meet the requirements of data timeliness in emergency scenarios.With the strenghths of high mobility,strong adaptability and swarms control,UAVs have obvious advantages for assisting wireless sensor networks to improve emergency data collection efficiency and backhaul capability.In emergency scenarios,it is critical for UAVs to quickly respond to obtain the optimal flight path and build a real-time data backhaul trajectory.However,the existing algorithms are difficult to solve efficiently and quickly for the problems of cooperative path planning of multiple UAVs and the construction of emergency data backhaul trajectory under the dynamic change of UAV topology.Therefore,this thesis takes the low-complexity and distributed characteristics of game theory into cosideration,and proposes a fast-response UAV path planning and real-time emergency data backhaul strategy to optimize the timeliness of emergency data with the help of coalition graph game and matching game theory for the multi-UAV-assisted wireless sensor network emergency data collection scenario.The main work and contributions of the thesis are as follows:In the scenario where the wireless sensor network convergence nodes are damaged after disaster,a multi-UAV path planning method based on hierarchical coalitional graph game is proposed to guarantee emergency data timeliness and assist data collection with multi-UAV.First,considering that the sensor node cluster and UAV swarms can improve the efficiency of emergency data collection,the UAV swarms data collection problem is decomposed into two sub-problems of node cluster and multi-UAV path planning.Then,the solution is respectively carried out using the exchange-based coalitional graph game and the improved Ant-Q algorithm-based coalitional graph game.Finally,compared with other existing algorithms,the simulation results show that the current method has faster convergence speed,which significantly reduces the age of information of emergency data and meet the requirements of data timeliness of collection in emergency scenarios.The problem of high age of information of the initially collected data in a large-scale sensing scenario can be effectively solved by constructing a real-time emergency data backhaul trajectory through UAV swarms.This thesis conducts a study on the multi-hop backhaul mechanism of emergency data,and proposes an emergency data backhaul method based on the multi-dimensional matching-based UAV swarms.First,the significant changes of UAV topology are captured using sampled dynamic graphs to obtain time-expanded graphs with update sequences,which can effectively reduce the space engaged by algorithm strategy.Second,the problem is decoupled into a multidimensional matching problem based on a loop hierarchical routing model,and the transmission direction of the emergency data is defined.Meanwhile,this thesis considers the data backhaul mode of opportunity transmission,which can further improve the timeliness of emergency data.Finally,according to the preference order of UAVs on the neighboring layer which is constructed based on the distance between UAVs and the amount of carried data,a multidimensional matching-based emergency data backhaul algorithm is proposed.Simulation results show that the method can effectively reduce the age of information of emergency data compared with existing algorithms.The thesis contains 25 figures,7 tables,and 110 references. |
PDF Full Text Request