A method of evaluating negative utility of information in presence of SSDF attack

The sensing security in cognitive radio networks (CRNs) draws people's attention day by day. Existing references mainly focus on detecting attacks, however the evaluation of negative utility caused by attacks has rarely been involved. Moreover, as a classical method of measuring information, Shannon information entropy and its derivations such as the weighted entropy (WE) and the utility information entropy (UIE) considered the utility of information, but they were limited to non-negative utility, neglecting the negative utility of information. In this paper, we propose the utility self-information and the real utility entropy of information (RUEI), which extends the range of information entropy from non-negative numbers to real numbers. The scenario of spectrum sensing data falsification (SSDF) attacks in CRN is analyzed to illustrate the application of RUEI, which shows that the RUEI can measure not only the non-negative utility of information from honest users, but also the negative utility of information from malicious users.