The Study On Related Problems Of Recovering Large-scale Coverage Holes In WSN By Multiple Airships Under Disaster Environment

As an emerging network technology, wireless sensor network(WSN) is regarded as a promising solution to monitor the physical world in unattended, unreachable, and even dangerous environments. But with the increasing of network size and the monitoring environment harming and complicating, the case that there are multiple large-scale coverage holes appeared simultaneously in WSN is inevitable. WSN tends therefore to be partitioned into multiple disjoint segments, which will result in that the damaged network topology can’t be perceived by top network. In this case, the normal network functions are destroyed, and the lifetime of WSN is ended. Hence, with the WSN having suffered natural disasters, how to recover the large-scale coverage holes in WSN has become a key issue to be solved. In this thesis, taking the airship as main air mobile node and facing multiple disjoint segments and large-scale coverage hole of the damaged WSN in a disaster environment, we establish multi-airship data gathering method and large-scale coverage hole boundary detection method for single disjoint segment in WSN to form the large-scale coverage hole boundaries of single disjoint segment. On the basis of that, we construct multi-airship simultaneous searching and federating method for multiple disjoint segments in WSN, thus forming the global large-scale coverage hole boundaries of damaged WSN and carring out the research on multi-airship large-scale coverage hole recovery method.Firstly, aiming at the disadvantages that the existing mobile data gathering approaches of WSN have low efficiency and high hardware overhead in the process of solving the data gathering of single disjoint segment, a dynamic clustering based multi-airships single disjoint segment data gathering method is proposed. According to the network model and its topological description, the definition for this data gathering problem is given, and its integer linear programming model is established with the optimization theory and dynamic clustering method. On the basis of that, a distributed heuristic data gathering algorithm is designed using the graph theory and the multi-robot task allocation strategy, and its effectiveness is verified through comparison of simulation results with experimental data.Secondly, aiming at the problem that the existing coverage hole boundary detection methods can not detect large-scale coverage hole boundary in a single disjoint segment quickly and efficiently, a boundary detection method for large-scale coverage holes in a single disjoint segment based on minimum critical threshold is proposed. The optimization problem of minimum critical threshold is highlighted, and its formulaic description is constructed according to probabilistic sensing model. On the basis of that, the distributed gradient information is used to approximately solve the optimization problem. With the proposed multi-airship single disjoint segment data gathering method, the local-scale rough boundary detection algorithm incorporating the minimum critical threshold and its iterative thinning algorithm are proposed according to blocking flow theory, and their execution efficiency and effectiveness are tested through theoretical analysis and comparison of simulation results with experimental data.Thirdly, aiming at the deficiencies that the existing federating method for disjoint segments in WSN is unable to adapt to the circumstance of considering the uncertainty in the segments distribution and the bandwidth-limited relaying node, a multiple airship simultaneous searching and bandwidth-aware relaying node federating method for disjoint segments in WSN is proposed. According to the related system model, the optimization problem for multiple airship simultaneous searching and bandwidth-aware relaying node federating is described, and it is decomposed into two sub-optimization problems, search-based airship deployment and relaying node deployment with selecting routing paths by using the iterative local search heuristic. On the basis of that, according to the proposed boundary detection method for large-scale coverage holes in a single disjoint segment, overlapping-aware connectivity search and bandwidth-aware clustered relaying node deployment are respectively proposed using the relevant theory of spectral graph and network flow, and their effectiveness are verified through comparison of simulation results with experimental data.Fourthly, aiming at the problem that the existing relay node deployment based coverage hole recovery method has low efficiency in the process of recovering the large-scale coverage hole in WSN, a homology-based large-scale coverage hole recovery method in WSN by multi-airship is proposed from the angle of message overhead and distribution uniformity for solving the position of optimal relay node deployment. For the remaining large-scale coverage hole after multiple disjoint segments in WSN have been federated, the positions of candidate relay nodes are incrementally generated by the Halton Sequence based on the homology theory. On the basis of that, the definition of degree and index value of simplicial complex is introduced, the redundant nodes in candidate relay nodes are eliminated by the related properties of the Cech complex, and then the positions of optimal relay node deployment are obtained. Through comparison of simulation results with experimental data, the effectiveness of proposed approach is verified.