Comparison between eigenvalue fusion and decision fusion for spectrum sensing of OFDMA signals under errors in the control channel

Recently, an eigenvalue fusion approach for detecting idle subchannels of OFDMA signals in the context of centralized cooperative spectrum sensing for cognitive radio (CR) was proposed. Four detection techniques were analyzed, and it was concluded that the eigenvalue fusion outperforms the decision fusion scheme, in spite of the larger volume of data sent to the fusion center (FC) in the eigenvalue fusion. Nevertheless, it was conjectured that bit errors in the reporting channel could be more disastrous to the data representing CR decisions than to the data carrying eigenvalues, masking a potential advantage of the eigenvalue combining also in terms of the volume of data sent to the FC. In this paper we investigate this conjecture and conclude that it is partially true: CR decisions are indeed more sensitive to channel errors, but the amount of redundancy inserted to protect the decisions so as to equate the performances of the two fusion schemes does not always leads to a larger number of bits in the decision fusion. Then, one needs to trade performance and amount of data in the reporting channel to decide upon which fusion scheme must be adopted, in a case by case analysis.