Superposed Band Detection Based on Error Probability Using Initial Likelihood Masking

Abstract

Superposed multicarrier transmission scheme is known to improve frequency utilization efficiency where several systems share spectrum without any spectrum spreading techniques. In superposed multicarrier transmission scheme forward error correction (FEC) coding is applied over subcarriers so that the effects of interference caused by superposition is mitigated by interleaving effects. When the interference power is large, the effects of FEC become smaller particularly owing to the mis-setting of initial log likelihood ratio (LLR). To solve this problem, FEC metric masking is proposed where initial LLRs for superposed subcarriers are replaced by neutral value in the receiver followed by decoding. This technique is effective, however, it needs the information about superposed frequency band. In this paper we propose a scheme to detect superposed band by using FEC metric masking. In the proposed scheme, the subcarriers where FEC metric masking is applied are changed and packet error rate (PER) is measured. When the FEC metric masking is applied to the superposed subcarriers, PER is minimized. Based on the change of PER, we can detect the superposed band. We evaluate the proposed scheme by computer simulation and show that it can detect the superposed band with high probability.