Link Adaptation and Resource Allocation in 5G Ultra Reliable Low Latency Communications

Abstract

Ultra-reliable and low-latency communications (URLLC) is a service category in fifth newlinegeneration (5G) wireless systems, which can support applications requiring very stringent reliability, latency, and availability. Achieving high reliability is challenging due newlineto various channel impairments like shadowing, interference etc. Macro-diversity, newlinewhere each user equipment (UE) is served by multiple base stations (BSs) is a proven newlinetechnique to achieve high reliability. 5G new radio (NR) enablers like flexible numerology, flexible frame structure, and short packet communication (SPC) can provide low newlinelatency. Resource allocation for macro-diversity enabled 5G URLLC is challenging newlinedue to reliability-latency trade-off, high resource usage, and exhaustive search required newlineto solve the resource allocation problem involving SPC. This thesis focuses on novel resource allocation techniques for macro-diversity schemes, viz. coordinated multi-point newline(CoMP), packet duplication, and maximal ratio combining (MRC), that can assure newlinethe required quality of service (QoS), viz. reliability, latency, and availability, to the newlineURLLC UEs, within the constraints of available resources. newlineInvestigative studies are required for evaluating and understanding the significance newlineof the various enablers for URLLC. The thesis starts with an investigative simulation newlinestudy of 5G URLLC. BSs selected to serve each UE based on the last reported reference newlinesignal received power (RSRP) might be outdated by the time of data transmission, newlinedue to the highly dynamic nature of the radio environment when millimeter wave newline(mmwave) spectrum is used. We extend the investigative study to evaluate the capability newlineof machine learning (ML) techniques in performing BS selection. newlineThere exist some challenges in using CoMP for URLLC in a cloud radio access newlinenetwork (C-RAN) architecture; mainly, fronthaul capacity and remote radio head newline(RRH) resource availability constraints.