I've used BPSK for this experiment. I've noticed that when the receiver is desynchronized in frequency with $\delta f=0.5T^{-1}$ (such that $T$ is the sampling frequency of the emitted signal), received constellation is scattered over a regular grid ?

Why this grid pattern occurs in this desync setup ?

I know that when phase shift (between emitter and receiver), then constellation rotates. If frequency shift (between emitter and receiver) , constellation rotates over time, so symbols are circularly spreaded. But in specific case of frequency shift equal to $0.5T^{-1}$ this circular spreading is not obvious anymore and this regular grid pattern appears. Symbols are caught in the way that cluster of symbols emerge and create this regular grid ?

Any ideas ?

Thank you

I've used BPSK for this experiment. I've noticed that when the receiver is desynchronized in frequency with $\delta f=0.5T^{-1}$ (such that $T$ is the sampling period of the emitted signal), received constellation is scattered over a regular grid ?

Why this grid pattern occurs in this desync setup ?

I know that when phase shift (between emitter and receiver), then constellation rotates. If frequency shift (between emitter and receiver) , constellation rotates over time, so symbols are circularly spreaded. But in specific case of frequency shift equal to $0.5T^{-1}$ this circular spreading is not obvious anymore and this regular grid pattern appears. Symbols are caught in the way that cluster of symbols emerge and create this regular grid ?

Any ideas ?

Thank you