Self-localization algorithm for deep mine Wireless Sensor Networks based on MDS and rigid subset

In order to adapt to the narrow space and complex branch in deep mine roadway, which have an adverse impact on the accuracy of nodes localization, a self-localization algorithm for deep mine wireless sensor networks based on MDS-MDS and rigid subset (Rigid-MDS) was proposed. The new distributed algorithm divides whole network into some globally rigid subset based on rigid graph theory, in which nodes have close spatial correlation, less hop and alike path direction with all of other nodes in the same subset, and it reduces the error by shortest path. And afterwards, some boundary nodes close to other subsets is chosen as framework with anchor nodes to locate the whole network. Finally, homogeneous coordinate system is used to make the geometric change. Simulation results show that the Rigid-MDS can reduce error in the roadway branch compared with MDS-MAP.