The Research And Realization On Sensor Localization Babed On Multidimensional Scaling

Wireless sensor networks, a novel technology about acquiring and processing information, have been applied to a wide variety of applications. Self-localization, that is, determing where a given node is physically or relatively located in the networks, is crucial for many applications. Multidimensional scaling (MDS) is a data analysis technique that transforms proximity information into a geometric embedding. MDS is well-suited to node localization in wireless networks, where the task is to use the correlative informations of nodes to determine the coordinates of nodes in a 2D or 3D space.A NMDS-RSSI (Nonmetric MDS - Received Signal Strength Indication) localization algorithm is presented. Nonmetric MDS is another type of MDS, where the proximity data of objects are only assumed to be related to distances by some monotone transformation. Ideally, the ranks of the nodes based on RSSI should be monotonic with their ranks based on true distance. The NMDS-RSSI approach runs nonmetric MDS on RSSI values of nodes instead of distances to localization, it can avoid the process of transforming RSSI into distance by other localization approaches based on RSSI, and can reduce errors about which this kind of transformation brings.In the real world, random RSSI fluctuations are present due to multi-path fading and shadowing. However, the inherent redundancy of nonmetric MDS makes NMDS-RSSI algorithm show robust to RSSI measurement errors in our experiments.Two NMDS-RSSI localization methods are presented: a centralized method that builds a global map and a distributed one that builds small local maps and then patches them together to from a global map. The distributed methods can suitable for large-scale networks.