TY - JOUR
T1 - Cross Validation Aided Approximated Message Passing Algorithm for User Identification in mMTC
AU - Yang, Linjie
AU - Fan, Pingzhi
AU - Li, Li
AU - Ding, Zhiguo
AU - Hao, Li
N1 - Funding Information:
Manuscript received February 8, 2021; accepted February 28, 2021. Date of publication March 4, 2021; date of current version June 10, 2021. This work was supported by National Key R&D Program of China (No.2018YFB1801104), NSFC Project (No.62020106001), and the 111 Project (No.111-2-14). The work of Li Li was supported by Sichuan Science and Technology Program (No.2020YFH0011). The associate editor coordinating the review of this letter and approving it for publication was D. Darsena. (Corresponding author: Li Li.) Linjie Yang, Pingzhi Fan, Li Li, and Li Hao are with the School of Information Science and Technology, Southwest Jiaotong University, Chengdu 610031, China (e-mail: [email protected]; [email protected]; [email protected]; [email protected]).
Publisher Copyright:
© 1997-2012 IEEE.
PY - 2021/6
Y1 - 2021/6
N2 - Recently, the approximate message passing (AMP) algorithm is advocated for overcoming the active user identification challenge in grant free (GF) massive random access (RA). However, requiring the perfect knowledge of user sparsity is a quite strong assumption of AMP algorithm in practical applications. Accordingly, we incorporate the cross validation (CV) method into the minimum mean square error (MMSE) denoiser based AMP algorithm, resulting a basic CV aided AMP algorithm (BCV-AMP) that can adaptively estimate the user sparsity. In order to further reduce its complexity, an improved CV aided MMSE denoiser based AMP algorithm (ICV-AMP) is proposed. Simulation results show that both the BCV-AMP and the ICV-AMP algorithms achieve almost the same user identification performance as the well-known MMSE denoiser based AMP algorithm, while the ICV-AMP algorithm has essentially the same complexity as that of the latter.
AB - Recently, the approximate message passing (AMP) algorithm is advocated for overcoming the active user identification challenge in grant free (GF) massive random access (RA). However, requiring the perfect knowledge of user sparsity is a quite strong assumption of AMP algorithm in practical applications. Accordingly, we incorporate the cross validation (CV) method into the minimum mean square error (MMSE) denoiser based AMP algorithm, resulting a basic CV aided AMP algorithm (BCV-AMP) that can adaptively estimate the user sparsity. In order to further reduce its complexity, an improved CV aided MMSE denoiser based AMP algorithm (ICV-AMP) is proposed. Simulation results show that both the BCV-AMP and the ICV-AMP algorithms achieve almost the same user identification performance as the well-known MMSE denoiser based AMP algorithm, while the ICV-AMP algorithm has essentially the same complexity as that of the latter.
KW - AMP algorithm
KW - compressed sensing
KW - cross-validation
KW - Grant-free
UR - http://www.scopus.com/inward/record.url?scp=85102261815&partnerID=8YFLogxK
U2 - 10.1109/LCOMM.2021.3064025
DO - 10.1109/LCOMM.2021.3064025
M3 - Article
AN - SCOPUS:85102261815
SN - 1089-7798
VL - 25
SP - 2077
EP - 2081
JO - IEEE Communications Letters
JF - IEEE Communications Letters
IS - 6
M1 - 9370106
ER -