Spectrum sensing is a key enabling functionality in cognitive radio (CR) networks, where the CRs act as secondary users that opportunistically access free frequency bands. Scattered spectrum sensing (SSS) enables a Cognitive Radio (CR) network to reliably detect licensed users and avoid causing interference to licensed communications. The data fusion technique is a key component of SSS. Byzantine failure problem with respect to data fusion is discussed, which may be caused by either malfunctioning sensing terminals or Spectrum Sensing Statistics Erroneous attacks (SSSE). In this paper various data fusion techniques will be investigated focusing on their robustness against Byzantine failures. Existing data fusion techniques use a fixed number of samples a new technique. A new technique is proposed that uses a variable number of samples. Evaluation of the proposed technique is made by comparing it with a variety of data fusion techniques in a network operating with multiple nodes conditions. Simulation results indicate that proposed technique performed better than against the Byzantine failure problem among the existing the data fusion techniques.
