We propose to reconstruct the IPNCM according to its definition and estimate the corresponding parameters, namely the SVs and powers of all interferences.
We introduce the idea of the subspace method to reduce the effect of the approximation error of using summation instead of the integral during the IPNCM reconstruction and obtain the SV of each interference with high enough accuracy.
Through the subspace method, we reconstruct a more precise IPNCM matrix with a very high correlation coefficient with the theoretical IPNCM.
We obtain a new SV estimation of SOI which is independent from the reconstructed IPNCM.