.png){kind=icon}
Return
Model-Driven Deep Unfolding LUAMP for Continuous Phase Modulation Equalization
.png){kind=icon}
PRE
AI
DOI：10.1109/lcomm.2026.3740533.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Continuous phase modulation (CPM) is attractive for telemetry communications because of its constant-envelope property, continuous phase, and high spectral efficiency. However, equalization of CPM over multipath fading channels remains challenging due to its nonlinear and memory-dependent nature, and existing equalizers still struggle to achieve a favorable balance among performance, robustness, and complexity. To address this issue, this letter proposes a model-driven, learning-based unitary approximate message passing (LUAMP) equalizer for CPM by deep unfolding and introducing a small number of trainable parameters. In addition, since the conventional mean-square-error objective cannot explicitly characterize the phase structure of CPM, a phase-aware loss is adopted, and a constant-envelope constraint term is further introduced to improve performance. Simulation results demonstrate that the proposed method achieves better BER performance than conventional linear and iterative equalizers over different multipath fading channels, while also bringing lower complexity. Under a longer pulse length, the proposed method achieves an SNR gain of about 4 dB over UAMP.
Keywords:
Continuous phase modulation (CPM)
Channel equalization
deep unfolding
LUAMP
Journal
IF:
4.4
Papers:
1.3W
Citations:
2.2W
Organization
Cited Papers
No cited papers available