Security-Oriented Pilot and Data Transmission for URLLC in Mission-Critical IoT Scenarios.

In this article, we focus on the joint design of channel training and data transmission for secure ultrareliable and low-latency communications (URLLCs) in mission-critical Internet of Things (IoT) scenarios, e.g., intelligent transportation and remote control. Specifically, we consider a multiple-input multiple-output multiantenna eavesdropper (MIMOME) system, and the role of artificial noise (AN) for securing URLLC in this system is studied. In the channel training phase, we use the two-way discriminatory channel estimation (DCE) protocol with AN injection to suppress the channel estimation accuracy at eavesdropper. Meanwhile, the AN-assisted secrecy beamforming scheme is adopted to mask the confidential data signals. Following by the security enhancements above, we provide a quantitative definition of achievable effective secrecy rate (AESR) to measure the performance of our URLLC system with imperfect channel state information (CSI) and short-packet feature. Then, a nonasymptotic closed-form lower bound of AESR is provided to make the numerical calculation tractable, and the asymptotic system performance in the high-SNR regime is also studied to gain a comprehensive insight. Based on the cyclic coordinated search method, we propose an iterative resource allocation algorithm to maximize the AESR of our system, where the blocklength and transmit power assigned to the reverse/forward pilots, confidential data signals, and AN are jointly optimized. In addition, numerical results reveal the AESR performance of our URLLC system for different system parameters, and demonstrate the convergence and superiority of our proposed algorithm.