Iterative frequency offset estimation based on singular value decomposition

This paper proposes an iterative frequency offset estimation based on singular value decomposition (SVD). The proposed technique achieves high resolution estimation of frequency offset even in the presence of strong co-channel and inter-symbol interference. Moreover, the proposed technique can estimate large frequency offset, e.g., ΔfT = 12.8 with high precision. The proposed technique comprises two new ideas. One is frequency offset estimation by using SVD and an adaptive filter. The other is iteration of the above frequency offset estimation followed by a frequency offset reduction. The iteration of the frequency offset estimation improves the estimation performance. In fact, it is evaluated by computer simulation that estimate the frequency offset normalized by a symbol duration, ΔfT = 12.8, in wireless sensor networks with direct sequence spread spectrum (DS-SS).