中文摘要：

极小BT参数下的GMSK信号解调面临着严重的码间串扰，最大似然序列检测（MLSE）算法的复杂度高，不适应于工程实践。给出了两种基于Laurent分解的低复杂度相干解调算法，即Wiener均衡器级联类OQPSK解调和减状态数Viterbi检测，均可以有效的克服码间串扰。仿真结果表明，相比MLSE检测，BT=0．3时，上述两种低复杂度解调算法的信噪比损失均小于0．5dB；BT=0．15时，减状态数Viterbi检测算法的信噪比损失小于0．5dB，而类OQPSK解调算法的信噪比损失大于2dB。

英文摘要：

Demodulation of GMSK signals with extremely low BT now faces fatal inter-symbol interference. Thealgorithm based on MLSE （Maximum Likelihood Sequence Detection）, for its high complexity, is un- suitable fur engineering practice. Two coherent demodulation algorithms based on Laurent＇s decomposition and with low complexity are given andstudied, i.e. OQPSK demodulation cascaded with Wiener equalizer and Viterbi Detection in reduced status. Both of the two algorithms can effectively overcome the inter-sym- bol interference. Simulation results show that, as compared with MLSE algorithm, the SNR losses of the two algorithms are less than 0.5 dB when BT is 0.3 . With lower BT （0.15） ,the SNR loss of Viterbi algo- rithm in reduced status is still less than O. 5 dB, while the SNR loss of OQPSK demodulation cascaded with Wiener equalizer is larger than 2 dB.