Millimeter-Wave Base Station Diversity and Human Blockage in Dense Urban Environments for Coordinated Multipoint (CoMP) Applications

摘要

The demand for mobile wireless access has seen an exponential rise in recent years and is expected to increase further as virtual-reality (VR) and augmented-reality (AR) applications become omnipresent. The growing demand has crowded operation in the sub-6 GHz bands typically used for mobile wireless access, which has motived the use of higher frequencies and wider bandwidths for future devices and data-driven applications. Millimeter-waves (mmWave) are vast and relatively underutilized and are an attractive option for meeting the extreme data demands for fifth-generation (5G) wireless systems. Fundamental propagation characteristics of the carrier frequencies in various operating environments are vital for producing realistic channel models that engineers can faithfully use for the design and simulation of future networks. At mmWave, channel characteristics are much different than at microwave frequencies, and thus legacy measurements and channel models are not adequate as a substitute for mmWave bands. Therefore, new measurements, channel models, and system analyses are required at mmWaves and for much wider bandwidths than traditional cellular systems in order to successfully deploy 5G mmWave networks.;This dissertation describes the necessary requirements for a mmWave channel sounder in order to accurately develop channel models for mmWave bands. Various mmWave channel sounder architectures and their specifications are compared, which is useful when designing a flexible and robust system. The dissertation presents a novel ultrawideband spread spectrum mmWave channel sounder that supports both a wideband sliding correlation mode and a real-time spread spectrum mode, also known as wideband correlation or direct correlation. Using propagation theory, several methods are presented herein to calibrate and verify the accuracy of the channel sounder. The sliding correlator configuration facilitates long-distance measurements with angular spread and delay spread for up to 185 dB of maximum measurable path loss, and the real-time spread spectrum mode is capable of supporting short-range and small-scale temporal measurements, which are vital for mmWave channel characterization.;This dissertation also presents two human blockage measurement campaigns with the real-time channel sounder that are used to characterize and model human blockage events at mmWave with directional antennas. First, a simple double knife-edge diffraction (DKED) model that accounts for antenna directivity is developed and is shown to accurately predict attenuation caused by human blockers when compared to measurements. Next, outdoor peer-to-peer measurements were conducted along a busy walkway to capture rapid fading events caused by pedestrians. The measurements are used to develop a four-state Markov model that is useful for simulating real-world dynamic fading events when developing network and higher-layer algorithms for future 5G mmWave systems.;A large-scale mmWave base station diversity measurement campaign at 73 GHz is also presented for which urban microcell (UMi) path loss models are developed, and multipath dispersion is characterized. Hypothesis testing with cross-validation is used to reveal that large-scale shadow fading of directional and omnidirectional path loss at an RX from nearest neighbor base stations can be modeled independently. Furthermore, simulated human blockage traces are superimposed on the base station diversity measurements to show that multiple base stations serving a user can reduce outage in the presence of rapid fading events compared to a single serving base station. Specifically, the percentage of users that experience outage with a -5 dB SNR threshold is shown to reduce from 24.7% when served by one base station, to 12.2% and 6.3% when served by two and three base stations, respectively, when the user employs selection diversity. Thus, network coordination and transmit diversity with directional beams at mmWave is shown to be useful when a user experiences signal degradation due to rapid fading events caused by humans.;This dissertation concludes by examining downlink precoding techniques in a 2x2 coordinated multipoint (CoMP) setting by utilizing the base station diversity measurements. Results show that when both users in a network experience interference, performance gains in network spectral efficiency are achievable by 81% of the networks that employ coordinated downlink precoding, and 90% of the networks achieve spectral efficiencies of 9.3 b/s/Hz or less, compared to 8.1 b/s/Hz or less for uncoordinated networks. While coordination is shown to improve network performance by suppressing interference when it exists, nearly half of the 680,000 CoMP networks (~43%) resulted in no interference for either user, meaning that CoMP may not be useful for interference coordination at mmWave with narrow directional beams. Therefore, mmWave UMi networks may not need to expend vast network resources for interference mitigation, since users and networks are shown to have marginal improvement when implementing downlink CoMP precoding due to the sparse existence of interference when employing narrowbeam and directional antennas at the base station and user.