With the development of the space-ground integrated information network, the Low Earth Orbit (LEO) satellite communication system is ushering in a development boom. The Paired Carrier Multiple Access (PCMA) technology is gradually developing to the low-orbit satellite communication due to its advantages of saving bandwidth resources. However, traditional PCMA technology is mostly used in high-orbit satellites, and cannot adapt to the highly dynamic fading channel character-istics of low-orbit satellite channels, which greatly degrades the bit error performance of the PCMA receiver. The bottleneck lies in the channel estimation and equalization technologies for overlapping signals.

Aiming at the channel characteristics of LEO satellites, this paper proposes a channel estimation scheme that combines training sequence estimation and Autoregressive (AR) model prediction. Based on the idea of superimposed training sequence channel estimation, an iterative method suitable for PCMA mixed signal channel estimation is introduced to improve the accuracy of training sequence channel estimation through iteration. The AR model is used to predict the Channel State Information (CSI) of the data sequence in real time. The use of AR model can also reduce the frequency of channel estimation in training sequences, so as to adapt to the dynamics of LEO satellite channels.

The simulation results show that the idea of superimposed training sequence channel estimation can be applied to the PCMA signals, and accurate channel estimation can be obtained after iterations. The method proposed in this paper can effectively improve the accuracy of channel estimation. After signal separation and demodulation, the bit error rate can reach the order of 10^-3when the signal-to-noise ratio is greater than 9 dB.

A channel estimation method for PCMA signal is proposed in this paper. The simulation results show that the bit error rate loss is within an acceptable range, which can support the application of PCMA technology in low-orbit satellite communication. The proposed algorithm has the advantages of simple structure, low complexity, and high practical value.