Tensor decomposition based-joint active device detection and channel estimation under frequency offset
In the future wireless cellular networks, the massive access supporting of the Internet of things (IoT) and machine type communication (MTC) has gradually become a pivotal requirement. To reduce collisions and signaling overhead generated by devices during access, grant-free random access (GF-RA) me...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | Chinese |
| Published: |
China InfoCom Media Group
2025-06-01
|
| Series: | 物联网学报 |
| Subjects: | |
| Online Access: | http://www.wlwxb.com.cn/zh/article/doi/10.11959/j.issn.2096-3750.2025.00424/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In the future wireless cellular networks, the massive access supporting of the Internet of things (IoT) and machine type communication (MTC) has gradually become a pivotal requirement. To reduce collisions and signaling overhead generated by devices during access, grant-free random access (GF-RA) method has been proposed. In GF-RA, the critical task is the joint active device detection and channel estimation (JADCE). However, in practical scenarios, low-cost IoT devices are usually equipped with inexpensive crystal oscillators to reduce costs, thus the frequency offsets are inevitable and seriously degrade the JADCE performance. The sporadic activity pattern of the devices enables the JADCE to be formulated as a large-scale sparsity constraint problem. In order to avoid the non-convexity introduced by the nonlinear between the frequency offsets and the channels, firstly, tensor decomposition was used to model the received signal from the perspective of the preamble sequence, channel, and frequency offset. Then, the alternating least square (ALS) method was exploited to solve the decomposed subproblems in parallel, and the estimated values of device activities, channel response and frequency offsets could be obtained simultaneously. Moreover, in order to make the subproblem strictly convex, the proximal minimization (PM) method was used to add the regularization constraints, which improved the convergence and stability of the proposed JADCE algorithm. Finally, the detection performance of the proposed algorithm was evaluated based on the number of antennas and the length of the preamble sequence. The simulation results show that the proposed JADCE algorithm achieves a missed detection probability close to 1.0×10<sup>-3</sup> within the given range of antenna numbers and the preamble sequence length variation, and approaches the normalized mean square error (NMSE) of channel estimation to 1.0×10<sup>-6</sup>. Compared with the existing algorithms under the frequency offsets, the proposed algorithm has a significant improvement in detection performance. |
|---|---|
| ISSN: | 2096-3750 |