ANN-based Collaborative Sensor Calibration and GA-approach to Sensor Mutation Management

Abstract

Collaborative sensor data service is an emerging technology and it is beneficial to various applications including robotics, medicals, industry and military. Sensor collaborative analysis may improve technical issues such as verification of sensor data or reduction of data transmission. However, typical approaches to sensor collaboration are less satisfactory. It is in part because the procedures for sensor collaboration are pre-defined and less adaptive to dynamic changes in environment, or it is because the procedures are developed one time and deployed for many times. They are not satisfactorily adaptive to environmental changes, or are less efficient to work the sensors collaborative to cope with abrupt changes. To improve sensor collaboration adaptive to environment changes, this paper proposes a two-tier deep learning technique to enable sensor devices.