Enhancement of vertical services leveraging 5g and enabling technologies

  1. Nakimuli, Winnie
Dirigida por:
  1. Jaime José García Reinoso Director/a

Universidad de defensa: Universidad Carlos III de Madrid

Fecha de defensa: 17 de junio de 2022

Tribunal:
  1. Javier Vales Alonso Presidente
  2. Marco Gramaglia Secretario/a
  3. Elisa Rojas Sánchez Vocal

Tipo: Tesis

Resumen

The fifth generation of mobile networks, i.e., 5G, aims to disrupt mobile networking and comes embedded with many capabilities (at both the radio-access and core networks) to support human and industrial communications. The previous generations of mobile networking, i.e., 4G, 3G, and 2.5G, aimed to satisfy human communications by offering voice, SMS, and Internet data services. Every subsequent generation focused on improving data rates and providing better network coverage. On the other hand, 5G aims to improve data rates and provide ubiquitous network coverage similar to the previous generations while simultaneously supporting a new set of communications, i.e., industrial communications. These communications focus on providing low-latency connectivity and high bandwidth to facilitate data exchange between industrial devices and the cloud. As a result, 5G has garnered much attention from various vertical industries, industries that provide applications/use cases with specific network requirements and do not own any network infrastructure. These vertical industries to be supported by 5G have been classified into the following sectors: industry 4.0, smart cities, smart transport, smart tourism, media \& entertainment, and public safety, each characterized by unique network requirements. Besides, 5G use cases have been broadly grouped into three main categories as per their network requirements:(i) eMBB, (ii) URLLC, and (iii) mMTC. The eMBB group encompasses use cases requiring peak data rates, whereas the URLLC category covers use cases requiring very low latency and ultra-high reliability. The latter category, i.e., mMTC, facilitates massive connectivity to billions of low-powered IoT devices. To deploy such a diverse range of use cases over a shared physical network infrastructure, the 5G system leverages the Network Slicing concept proposed by the NGMN alliance. Network slicing creates complete and isolated logical networks (composed of virtualized instances of both RAN and CN functions) on top of shared physical infrastructure. These complete logical networks (also known as Network slices) are employed to cater to use cases with specific network requirements. Thus, the network slicing concept contributes greatly to the current flexibility, scalability, agility, and programmability attributes of the 5G network. On the other hand, this network slicing concept is greatly enabled by NFV and MEC technologies. The NFV technology decouples NFs such as DNS, routing, load balancing, and firewall from legacy software and hardware vendor equipment by implementing them as VNFs that can be deployed on general-purpose servers. This NFV approach provides flexible deployment of NFs, leveraging all the benefits of virtualization. Conversely, 5G, when enabled with the MEC technology, comes with the availability of networking, computing, storage, and real-time context information close to the edge of the network (i.e., within the RAN). Accordingly, vertical use cases with stringent computing, storage, and networking requirements can benefit significantly from these MEC resources and services. However, some of these envisioned vertical use cases are currently supported by proprietary or other wired and wireless networks. Moreover, for some verticals, their use case requirements are too stringent and therefore not satisfied by the current wired and wireless networks. Hence, this puts 5G, and its enabling technologies in a unique position to have to: (i) offer much better performance gains compared to the existing networks, (ii) satisfy the stringent network requirements currently not supported by existing networks, and (iii) increase and provide new revenue streams to all involved parties (i.e., users, verticals, mobile operators and IT \& Telecom vendors), in order to justify the shift to 5G given the associated costs. Nevertheless, 5G and its enabling technologies are well poised to satisfy these heterogeneous requirements and support various vertical industries. Moreover, these vertical industries play a significant factor in increasing the adoption of 5G. Hence, there is a great need for partnership and collaboration between the mobile operators offering the 5G technology and these vertical industries. Besides, most of these verticals have little or no knowledge about telecommunications and networking. Subsequently, there is a gap between verticals that are well-conversant with the use case functions \& capabilities and mobile operators who are the networking experts with no knowledge about the vertical use case functionalities. This thesis will be focused on analyzing the performance enhancements that 5G and its enabling technologies bring to the vertical industries arena by focusing on some vital 5G key performance indicators (KPIs). Consequently, we leverage some of the envisioned 5G vertical use cases affected by such KPIs to evaluate these technologies' improvements to such use case scenarios. To this end, we selected some of the primary 5G KPIs, i.e., service creation time, end-to-end latency, reliability, and user-experienced data rate. For each KPI, we leveraged one of the envisioned vertical use cases to evaluate the performance improvements that 5G and its enabling technologies bring to the use case. In the context of this thesis, the 5G service creation time KPI refers to the total time it takes to prepare, instantiate, configure \& activate, modify and terminate the vertical service. The end-to-end latency KPI is evaluated as the network delay experienced by the packet from the time the source node sends the packet to the time the destination node receives it and is also referred to as the OTT latency. This KPI is directly related to the RTT latency, which refers to the time taken to send a message from the transmitting network node to the time taken to receive an acknowledgment from the receiving network node. The reliability KPI refers to the ratio of the number of packets successfully transmitted to a given network node within a specified time constraint to the total number of network packets sent to the network node. This thesis measured the reliability KPI using the packet loss ratio. Besides, the user-experienced data rate KPI comprises the DL and UL data rate components. It refers to the minimum achievable data rate for a user in a particular direction (i.e., either DL or UL) in a real network environment. In addition, this thesis aims to bridge the gap between the mobile operator and the vertical industries by aiding verticals with little networking knowledge in translating their use case requirements in terms of network, storage, and computing resources into VNFs. These VNFs can be interconnected following the specified network topologies creating network services that fulfill the use case requirements. These network services can easily be deployed, removed, and scaled on-demand at the optimal place in the network to satisfy all requirements, leveraging 5G and 5G-enabled platforms and technologies. To this end, in this thesis, our research was focused on the following objectives: • A comprehensive analysis of the service creation time KPI and its associated phases is one of the 5G KPIs defined by 5G-PPP. • Validation of the 5G end-to-end latency and reliability KPIs. For these KPIs, our initial contribution was designing, implementing, monitoring, and validating a latency-sensitive vertical use case over a 5G-Enabled platform using a real 5G private network infrastructure. Subsequently, we examined the use case performance under varying delays and packet loss radio conditions. Consequently, we extended this contribution by proposing a modification to the vertical service that would improve the use case performance (in the presence of varying delays) by utilizing some of the services provided by the MEC technology. • Evaluation of the 5G user-experienced data rate and end-to-end latency KPIs. We leveraged an ultra-high bandwidth vertical use case implemented across multi-sites over a 5G-enabled platform for this contribution.