Browsing by Subject "5G"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    A Comprehensive Examination of Emerging 5G Services: Challenges and Opportunities
    (2021-09) Narayanan, Arvind
    2019 marks the year when 5G services were rolled out commercially to consumers. 5G is expected to support sub-millisecond latency as well as ultra-high throughput of 20~Gbps that is a 100x improvement compared to its predecessor 4G/LTE. However, there exists a vacuum in understanding how 5G as a technology performs in-the-wild and whether it can fulfill its promises. Even after two years of being deployed commercially, the impact of 5G services on network performance or power consumption is not well understood. It is also unclear how applications and services today practically benefit from 5G technology. In an attempt to fill these voids, this dissertation is dedicated to examine the commercial 5G landscape ``from'' and ``across'' several key dimensions -- network performance, radio power characteristics, and quality-of-experience implications for mobile applications -- to provide key insights for interested stakeholders, and propose intelligent mechanisms for application developers to better leverage 5G technology that can aid in building future 5G-aware apps and services. To that end, using tools such as 5G Tracker developed in-house, we conducted a detailed measurement study to provide a first impression of the network performance characteristics of 5G services (including that of the much anticipated mmWave) and helped establish an important baseline for studying the evolution of 5G's performance. Subsequently, we expanded its scope and conducted the first comprehensive measurement study to investigate power consumption of 5G radio on commodity smartphones. Overall, our findings reveal key characteristics of commercial 5G services in terms of throughput, latency, performance bottlenecks, coverage, radio state transitions and radio power consumption under diverse scenarios, application performance (web-browsing, file-download, video streaming) with detailed comparisons to 4G/LTE networks. Furthermore, we also quantitatively reveal critical trade-offs (e.g. ultra-high vs. stable network performance, performance vs. energy) that get amplified with 5G and how challenges lie for upper layers (transport and up) to effectively utilize 5G. Leveraging the insights obtained from our measurement studies, we statistically reveal that, unlike in 4G/LTE, location alone is not sufficient for mapping 5G (especially mmWave) performance. We further proposed Lumos5G, a data-driven machine learning based framework for mmWave 5G throughput prediction that utilizes and fuses a variety of contextual features that can be collected on the smartphone, including its location, mobility information, geometric relationship with the 5G base station, and cellular connectivity information for predicting 5G throughput. We discuss how approaches like these provide opportunities as well as challenges in building future 5G-aware apps. In summary, this thesis contributes to better understanding the emerging commercial 5G landscape in the U.S. by systematically designing measurement methodologies and timely conducting experiments and analysis. With the rate at which 5G advancements and adoption are taking place, clearly 5G is on its evolutionary path for the next few years. The insights and challenges highlighted by our research will hold a mirror up to the implications of this new emerging technology. Artifacts of our research have been made available in the form of: datasets (5Gophers v1.0, Lumos5G v1.0, SIGCOMM21-5G v1.0), a smartphone-based measurement tool (5G Tracker) and a 5G mapping platform (5Gophers.umn.edu) for visualizing 5G coverage.
  • Thumbnail Image
    Item
    Data Conversion Techniques for Next Generation Communications
    (2017-12) Saha, Anindya
    The voice-only mobile-telephony 1G systems have evolved a long way to today’s data-driven 4G LTE networks, causing an exponential increase in mobile broadband data consumption. Furthermore, 5G is expected to deliver unprecedented data rates (tens of Gbps) exploiting mm-wave bands (30-300 GHz). Analog-to-digital converters (ADC) are one of the crucial factors in determining the pursued data rates. In the first part of this dissertation, a 100MS/s 9-bit companding SAR ADC, which exploits the statistical properties to reduce the PAPR of broadband multi-carrier signals in 4G LTE has been investigated. The architecture provides amplitude-specific gain with a fast instantaneous AGC, reducing the effects of PAPR and optimizing quantization noise, emulating the performance of a higher resolution ADC. Additionally, gain-before-sampling results in reduced sampling capacitor size, which lowers power and area. In the second part of this dissertation, a 1 GS/s 7-bit ADC using PWM technique and time-domain quantization is investigated to harness the benefits of the rapidly improving time resolution, so that the envisioned data rates in 5G can be realized with the lowest possible power. Thanks to digital delay line based time-domain computations, proposed architecture is highly digital therefore scalable, which is beneficial since scaling does not favor voltage-domain circuits.
  • Thumbnail Image
    Item
    An Engineer's Journey into Network Function Virtualization & 5G Research
    (2021-05) Quant, Jacob
    Softwarization (using software to provide functionality previously performed by hardware) has been driving developments in the computer networking field for more than a decade. Two examples of this are SDN (software-defined networking) and NFV (network function virtualization). Both of these play important roles in ushering in new technologies such as 5G (5th generation standards for high-speed cellular networks).Ever-increasing NIC (network interface card) data transfer rates necessitate improvements in NFV system design in order to avoid degrading throughput. This thesis introduces NFlambda and summarizes my contributions to it as well as the 5G-Tracker project. NFlambda is an NFV framework designed to facilitate efficient scaling of virtual network functions (VNFs) so that they can operate at line rates in excess of 100Gbps using commodity hardware. To achieve this in practice (i.e. without artificiallycontriving the traffic profile or using unrealistically simple VNFs) it becomes essential to avoid the timing penalties imposed by having to access the last level cache (LLC) or main memory. NFlambda achieves this primarily by decomposing VNFs into finer-grained components, which can be scaled independently and, in many cases, avoid having to share their state among multiple instances (running on separate CPU cores). Several key contributions that I made to this work are: adding support for YAML-based configuration, developing a proof-of-concept protocol for integrating an external controller, automating experiment design & execution, and assisting with the implementation of an IPsec VNF. 5G-Tracker is a crowd-sourced system for collecting and analyzing data related to commercial 5G network deployments. It can be used to build coverage maps, identify contextual factors affecting performance, and more. My work on this project focused on the development and documentation of the API used by the mobile application to communicate with back-end servers and the design of a web interface to support collaboration among researchers using the platform.
  • Thumbnail Image
    Item
    Enhancing the Performance of Mobile Video Streaming Ecosystems
    (2022-12) Shehata, Eman
    Recent years have witnessed a rapid increase in video streaming services (e.g., Netflix, YouTube, Amazon Video, ... etc) to meet users' interests as a result of the massive content published by content providers, high-speed Internet, the wide use of social networks, along with the growth in smart mobile devices. Additionally, the recent deployment of commercial 5G in 2019 and its potential for ultra-high bandwidth has enabled a new era for bandwidth-intensive networked applications such as volumetric video streaming. This growth in available content and demand places a significant burden on the Internet infrastructure. In addition to the complex structure of videos as each video is encoded in multiple resolutions, and different bitrate quality levels to support diverse end-user devices and network conditions. Thus, large-scale content providers have resorted to employing one or more content distribution networks (CDNs) to cache video content and handle user requests, as well as resorting to edge computing and machine learning to improve the performance perceived by their end users. Poor performance impacts user engagement, which leads to significant revenue loss for content providers. In this thesis, we discuss crucial research problems to improve the performance of mobile video streaming ecosystems to meet the scalability and user QoE performance requirements. First, we study the performance of intermediate caches in a hierarchical cache network. We show that when cache servers at different layers act independently this leads to caching objects which are evicted before their next request arrives leading to cache under-utilization.To overcome this issue, we proposed "BIG" cache abstraction which deals with distributed cache pieces as if they are "glued" together to form one "virtual" "BIG" cache. Thus, allowing any existing caching strategy to be applied as a single consistent policy for this "BIG" Cache. Consequently, "BIG" cache improves object hit probability, thereby minimizing the origin server load, and network bandwidth. Second, object access patterns are frequently changing due to the frequent changes in object popularity due to its diurnal access pattern, and during its life span. Due to these frequent changes, caching algorithms cannot rely on the locally observed object access patterns for making caching decisions. On the other hand, manually tuning the caching algorithm for each cache server according to the changes in the request access patterns is very expensive and is not scalable. To address this issue, we developed a machine-learning LSTM Encoder-Decoder model for content popularity prediction. Our DEEPCACHE is a self-adaptive caching framework for making end-to-end caching decisions based on the predicted popularity. We show that it manages to increase the number of cache hits for existing caching policies. Third, routing is a central problem to ensure the resiliency of CDNs. Purely distributed routing algorithms such as Bellman-Ford suffer from the "count-to-infinity" problem, whereas Dijkstra's algorithm requires global topology dissemination and route recomputation. Much of the recent literature on resilient routing is resilient to k link/node failures for a constant k (and often placing topological constraints on the graphs), and none of them work under arbitrary link failures. To address this issue, we developed a proactive routing algorithm that ensures the connectivity between any pair of nodes under arbitrary failures without the need for global topology dissemination and route recomputation as in purely distributed routing algorithms. Our algorithm limits the number of nodes involved in the recovery process as well as the number of link reversals, and convergence time. An additional advantage is the ability to utilize multiple paths to send traffic between nodes due to utilizing directed edges between nodes even upon failures. Finally, with the recent deployment of commercial 5G in 2019 and its potential for ultra-high bandwidth, we studied the characteristics of 5G throughput and its impact on video streaming applications. Our findings show that the wild fluctuations in 5G throughput and its dead zones lead to a large stall time while streaming videos. We redesigned video streaming applications to be 5G-Aware taking full advantage of the ultra-high bandwidth and overcoming its varying throughput. Our experiments show that our proposed strategies consistently deliver high video quality close to the theoretical optimal results reducing (if not eliminating) the stall time.