Aref, Abed Doosti;
Zhu, Xu;
Wen, Miaowen;
Masouros, Christos;
Krikidis, Ioannis;
(2025)
OTSM With Delay-Doppler Alignment Modulation Meets mmWave mMIMO ISCAP: Waveform Optimization by Deep Reinforcement Learning.
IEEE Internet of Things Journal
10.1109/jiot.2025.3582031.
(In press).
Preview |
Text
OTSM_With_Delay-Doppler_Alignment_Modulation_Meets_mmWave_mMIMO_ISCAP_Waveform_Optimization_by_Deep_Reinforcement_Learning.pdf - Accepted Version Download (6MB) | Preview |
Abstract
Orthogonal time sequency multiplexing (OTSM) has arisen as a promising single-carrier waveform, providing reliable performance akin to orthogonal time frequency space (OTFS) while surpassing orthogonal frequency division multiplexing (OFDM) in high-mobility doubly spread channels (DSCs), yet with significantly lower complexity. To alleviate both time and frequency dispersions in DSCs, delay-Doppler alignment modulation (DDAM) has been recently proposed for the systems operating at millimeter-wave (mmWave) and higher frequency bands. Building on the promising combination of OTSM with DDAM, this paper presents for the first time, waveform optimization in mmWave massive multiple-input multiple-output (mMIMO) integrated sensing, communication, and wireless power transfer (ISCAP) systems. We propose an OTSM-DDAM-based ISCAP system and derive the system’s input-output relations in time, delay-time, and delay-sequency (DS) domains based on delay-Doppler (DD) bin alignment in the presence of fractional DD shifts and by incorporating transceiver hardware impairments. This facilitates the derivation of key performance metrics including bit error rate (BER), spectral efficiency (SE), Cramér-Rao bound (CRB) for sensing, and energy harvested via wireless power transfer (WPT). By harnessing the advantage actor-critic method combined with a greedy strategy, an on-policy deep reinforcement learning algorithm is introduced to solve a newly defined optimization problem. This approach optimizes the ISCAP waveform while simultaneously recognizing the environment, providing the base station with the path state information needed for practical DDAM implementation. The optimization problem jointly considers precoding, power loading, time and power splitting ratios, beam alignment, and receive combining with the goal of minimizing the transmitted power coupled with CRB while maintaining constraints on signal-to-interference-plus-noise ratio, BER, SE, CRB, harvested energy, and total power budget to ensure high-quality ISCAP services. Simulation results, based on both real-world and deep learning datasets, validate the effectiveness of the proposed scheme, demonstrating improvements in peak-to-average power ratio, SE, detection complexity, CRB, and BER in high-mobility DSCs.
| Type: | Article |
|---|---|
| Title: | OTSM With Delay-Doppler Alignment Modulation Meets mmWave mMIMO ISCAP: Waveform Optimization by Deep Reinforcement Learning |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1109/jiot.2025.3582031 |
| Publisher version: | https://doi.org/10.1109/jiot.2025.3582031 |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. |
| Keywords: | OTSM, DDAM, ISCAP, WPT, massive MIMO, mmWave, waveform optimization, deep reinforcement learning |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Electronic and Electrical Eng |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10211379 |
Archive Staff Only
 |
View Item |
