非连续正交频分复用（NC-OFDM）是认知无线电系统十分有前景的物理层传输方案，而准确的载波频偏和IQ失衡估计对其性能至关重要。目前针对OFDM系统的载波频偏和IQ失衡联合估计依赖训练符号，由于NC-OFDM的可用频谱随机分布且不连续，因此已有方法不能应用于NC-OFDM系统。盲算法不需要训练符号，从而提供了可能的解决方法。本项目采用恒定包络调制技术以及利用空载波，研究NC-OFDM系统的载波频偏和IQ失衡联合盲估计，算法适用于SISO和MIMO系统。在已有的研究基础上，项目通过恒定包络调制设计了适用于任意星座图的频偏估计方法，同时利用空载波在有频偏的情况下估计IQ失衡。我们还研究通过两步式迭代的估计方法以提高精度。针对盲算法的可识别性以及复杂度高的问题展开研究，通过曲线拟合提出了简化的算法。利用正交空时编码，算法可以应用在MIMO系统中，进一步增强普适性。

Non-continuous orthognal frequency division multiplexing (NC-OFDM) is a promising physical transmission technique in cognitive radio, and the carrier frequency offset (CFO) and IQ imbalance estimation is of great importance. Currently, the joint CFO and IQ imbalance estimation methods for conventional OFDM rely on the training symbols, thus they are not applicable in NC-OFDM systems for the available spectrum in NC-OFDM is random and non-continuous. The blind estimation provides a possible solution as it needs no training symbols. This project aims to study the joint blind estimation of CFO and IQ imbalance with the help of constant envelope modulation and null subcarriers, both SISO and MIMO systems are under consideration. Based on the exiting research, this project proposed a blind CFO estimation scheme suitable for arbitrary constellations via exploiting the constant envelope transform. The blind estimation of IQ imbalance is carried out in the presence of CFO. We designed a two-step mechanism to further improve the estimation accuracy. The inherent problems of identifiability and high complexity of blind algorithms are addressed, and a low complexity estimation is introduced by utilizing the curve fitting. With the application of orthogonal space time block codes, the proposed scheme can also be applicable in MIMO systems, which further increases the universality.