53results about How to "Improve Spectrum Sensing Performance" patented technology

The invention discloses an energy detection method based on weighting matrix filtering. The energy detection method based on weighting matrix filtering comprises the steps that sampling is conducted on signals of a monitored frequency band, so that sampled signals are obtained; signals of all sampling points of the sampled signals are grouped, and each group of signals of the sampling points form a column vector; a weighting matrix used for filtering is obtained through the convex optimization technology according to the frequency range of a monitored channel needing spectrum sensing and the total number of the signals of the sampling points of each group; the energy of the monitored channel needing spectrum sensing is worked out through the weighting matrix and all the column vectors; through comparison between the energy of the monitored channel and a judgment threshold, whether an authorized user signal exists in the monitored channel or not is judged, and spectrum sensing is achieved. The energy detection method based on weighting matrix filtering has the advantages that according to the obtained weighting matrix, the energy of the monitored channel can be directly obtained; in addition, when the order of the weighting matrix used for filtering is equal to the order of an FIR filter, higher spectrum sensing performance can be obtained.

The invention discloses an energy detection method for restraining direct current bias. A sensor node is used for performing sampling on signals of a monitoring channel, sampled signals are obtained, then, difference operation is performed on signals of two adjacent sampling points in the sampled signals, a differential signal of the signals of the two adjacent sampling points in the sampled signals is obtained, then the total energy of all the differential signals is calculated, finally, the total energy of all the differential signals is compared with a judgment threshold, whether a user is authorized with the signals in the monitoring channel to achieve spectrum sensing or not is judged, and the spectrum sensing performance is high. In addition, due to the fact that when the judgment threshold is calculated, only noise power and false alarm probability need to be known, power of the direct current bias does not need to be known, in this way, when the power of the direct current bias is increased, the spectrum sensing performance can not be reduced, but when the power of the direct current bias is larger than one fourth of the noise power, the method has the better spectrum sensing performance compared with an existing energy detection method.

The invention discloses an OFDM signal spectrum sensing method for time synchronization in the presence of frequency offset. The method comprises the following steps: sampling a received signal from a monitoring channel to obtain a sampling signal; then estimating noise power via the autocorrelation of a cyclic prefix of an OFDM signal in the received signal according to the sampling values of all sampling points in the sampling signal under the condition of time synchronization and in the presence of frequency offset; calculating a covariance matrix of the sampling signal according to the sampling values of all sampling points in the sampling signal; then calculating test statistics according to the noise power and the covariance matrix of the sampling signal; and finally, comparing the test statistics with an obtained judgment threshold to judge whether an authorized user signal exists in the monitoring channel so as to realize spectrum sensing. The OFDM signal spectrum sensing method has the advantages of effectively improving the spectrum sensing performance of the OFDM signal and ensuring low computation complexity.