HomePublicationsTechnical Report

Ultra-Reliable Low-Latency 5G Communications in Autonomous Vehicular Networks
Ref: CISTER-TR-191213       Publication Date: 13, Dec, 2019

Abstract:
Fifth generation network (5G) represents a paradigm shift in mobile and object (things) communications. 5G will propose new services that require large broadband data rates as well as ultra-reliable and low latency links. Some other features of 5G include heterogeneous networks with object densification, virtualized radio access, mm-wave transmission, and very high device densities. Unlike previous generations, advanced 5G concepts aim to integrate multiple radio technologies thanks to network slicing concepts (software defined radio and network technologies). New URLLC (ultra-reliable low latency communication) services are characterized by the need to support industrial real time communication, where successful data transmission must be guaranteed within low (deterministic) latency bounds in the order of 1ms for the radio interface and 3ms end-to-end. 5G is currently being standardized and tested as the new radio (NR) technology by the 3GPP (Third generation Partnership Project). Autonomous vehicles have emerged as a main trend in vehicle development over the next decade. To support autonomous vehicles, ultra-reliable low-latency communication (URLLC) is required between vehicles and infrastructure networks, e.g., a fifth generation (5G) cellular networks. Hence, reliability and latency must be jointly investigated in 5G autonomous vehicular networks. Vehicle platoon is an advanced form of autonomous vehicle organization and optimization for traffic flows in dense urban environments. Vehicle platoons can reduce complexity for autonomous decisions in dense vehicular environments and can also improve connectivity with infrastructure by providing more points of communication with the vehicular fog/cloud/edge infrastructure. In this research, we will explore the problem of ultra-reliable and low-latency communication in millimeter wave-enabled massive multiple-input multiple-output (MIMO) networks for V2X applications. The problem is formulated for vehicular platoon formations, exploring the use of 5G as an enabler of ultra-reliable and low latency communication for vehicle to vehicle and vehicle to infrastructure communications. Different assumptions will be considered for each of the involved nodes with conventional or the new massive MIMO antenna technology. Aspects such as 3D-beamorming, interference cancellation, non-orthogonal multiple access (NOMA), modified OFDM (Orthogonal Frequency Division Multiplexing) modulation such as FBMC (Filter Bank Modulation and Coding), and retransmission-spatial diversity will be explored. Novel methods to evaluate latency in vehicular platoon formations will also be investigated. Novel channel assumptions such as asymmetric correlation, non-stationary multipath and multi-ray channel models will be considered. In this study, we propose a reliability and latency joint function to evaluate the joint impact of these two metrics in 5G autonomous vehicular networks. The interactions between reliability and latency are illustrated via simulations of 5G autonomous vehicular networks. System level scenarios with Manhattan grid networks will be used to emulate large urban deployments with hundreds or thousands of vehicle platoons, interferers, detailed propagation models and interactions between the fixed roadside infrastructure and the vehicular mobile nodes. A new solution that improve both the reliability and latency performance and ensure URLLC is presented for 5G autonomous vehicular networks. Finally, implement a set of promising algorithms in a software defined radio platform and evaluate at the system level a full version of the proposed algorithms.

Amir Hossein Farzamiyan