Data recovery algorithm under intrusion attack for energy internet

摘要

The openness, sharing and interaction of data pose many challenges to the data security protection of the Energy Internet. Most of the existing methods focus on data encryption and reliable data storage. However, these methods cannot avoid attacks on data in the Energy Internet. To solve data security under the condition that the data is attacked, this paper addresses optimal data partition and recovery considering attacks for the Energy Internet. Firstly, in order to greatly reduce data dimension, we propose an attribute reduction algorithm based on indiscernibility and equivalence (AR-IE) and an optimal data partition algorithm via attribute reduction Rough Set (ODP-ARRS). Secondly, on the basis of ODP-ARRS, this paper presents a data recovery algorithm by using the erasure code (DR-EC). Finally, three comparative experiments are implemented on the five standard datasets from UCI. Experimental results show that AR-IE algorithm has the obvious effect of attribute reduction for experimental datasets so that the average number of attributes is reduced by 64.27%, and the running time of ODP-ARRS and DR-EC algorithm has advantages in terms of running time. (C) 2019 Elsevier B.V. All rights reserved.