The Internet of things (IoT) is the new paradigm which extends the traditional Internet from laptops and mobile phones to different set of devices with own processing and communicating capability. Mobility among IoT devices challenges devices energy efficiency as these devices are battery powered. Most of the existing work lacks to provide efficient connectivity when devices are mobile. Hence, in this work, proactive mobility support for IoT devices has been propounded. Here, Kalman filter is used to predict the point of attachment of a mobile node and trajectory. Fault tolerant characteristics are being considered in the proposed algorithm to prevent the network from failing and for efficient routing of data. Proposed approach performs well in terms of handoff delay, movement prediction and prediction accuracy when compared to existing approach.