Abstract:Millimeter-wave (mmWave) grant-free (GF) access is a promising approach for massive machine-type communication (mMTC) networks to improve the access efficiency and alleviate the shortage of spectrum resources. Due to the lack of authentication, mmWave GF-mMTC networks are vulnerable to the pilot contamination attack (PCA), which can cause severe performance degradation of the channel estimation (CE) and user activity identification (UAI). However, the existing PCA resistance schemes for mmWave GF-mMTC networks perform the CE, UAI, and PCA detection through two separated phases, which will limit the system performance. To solve the problem, we establish a three-dimensional (3-D) transmission model with time-correlated two-dimensional sparsity for mmWave GF-mMTC networks under PCA, where the user activity sparsity, the virtual angular channel sparsity, and the temporal correlation of legitimate user (LU) status are jointly considered. Based on the established transmission model, we develop a 3-D compressive sensing based joint CE, UAI, and PCA detection (3D-CS-JCUPD) scheme. In this scheme, a parallel expectation-maximization vector approximate message passing with multiple measurement vector (Parallel EM-VAMP-MMV) algorithm is proposed to estimate the channel virtual representation (CVR) and the LU status is identified with the aid of different temporal correlation features between LUs and attackers. Moreover, we also develop a location information aided joint CE, UAI, and PCA detection (LIA-JCUPD) scheme to address the situation when attackers and LUs exhibit similar temporal correlations, where the BS utilizes the recorded LU location information to distinguish the LU status. Simulation results show that the developed schemes can achieve substantial performance gains over several reference schemes.

Scalable Active User Detection Based on Federated Learning in Cell-Free Massive MIMO System

A Scalable Deep-Learning-Based Active User Detection Approach for SEU-Assisted Cell-Free Massive MIMO Systems

Adaptive Federated Learning-based Joint Pilot Design and Active User Detection in Scalable Cell-free Massive MIMO Systems

Active Detection and Channel Estimation Schemes for Massive Random Access in User-Centric Cell-Free Massive MIMO System

Joint Active User Detection, Channel Estimation, and Data Detection for Massive Grant-Free Transmission in Cell-Free Systems

Deep Learning Based Active User Detection for Uplink Grant-Free Access

Asynchronous Grant-Free Random Access: Partially Uni-Directional Message Passing for Joint User Activity Detection and Channel Estimation.

Joint Channel Estimation and Active User Detection for Cell-Free Massive Access System Exploiting Coarse User Location Information

Efficient Hybrid Distributed Detectors for Cell-Free Massive MIMO Systems

Federated Learning with Unsourced Random Access.

A Flexible Framework for Grant-Free Random Access in Cell-Free Massive MIMO Systems

Enabling Efficient Scheduling Policy in Intelligent Reflecting Surface Aided Federated Learning

Joint Active User Detection and Channel Estimation in Massive Access Systems Exploiting Reed-Muller Sequences

Large-scale Fading Decoding Aided User-centric Cell-free Massive MIMO: Uplink Error Probability Analysis and Detector Design

Sensing User's Activity, Channel, and Location with Near-Field Extra-Large-Scale MIMO

Deep Learning-Based Approach for User Activity Detection with Grant-Free Random Access in Cell-Free Massive MIMO

EM-AMP-Based Joint Active User Detection and Channel Estimation in Cell-Free System

Joint Channel Estimation and User Activity Detection for Mmwave Grant-Free Massive MTC Networks under Pilot Contamination Attack

Data-aided Active User Detection with False Alarm Correction in Grant-Free Transmission

EP-based Joint Active User Detection and Channel Estimation for Massive Machine-Type Communications

Sparsity Learning Based Multiuser Detection in Grant-Free Massive-Device Multiple Access