Characterization and Measurement of Communication Channels in the mmWave Band

Abstract

5G wireless technology promises to change society as we know it. This new technology enhances the user experience by increasing the network data rate and the number of concurrent users, and reducing latency and power consumption. To ensure all these improvements, there is a technological leap where all layers of the network are involved. From a physical layer perspective, the main innovation is the frequency increase towards the millimeter wave band, also known as the mmWave band. To date, communication systems have been based on the sub-6 GHz spectrum, which is currently saturated due to the limited bandwidth. To address this issue, the research community is focusing on the feasibility of the mmWave band (30 GHz - 300 GHz) for commercial communications applications, since this band offers great advantages, such as large spectrum availability. However, there are also several technological challenges associated with millimeter waves, such as higher attenuation losses and the paradigm shift in propagation mechanisms compared to the sub-6 GHz band. In order to meet the necessary requirements of this band, it is therefore essential to perform a detailed study and characterization of the propagation channel involved in the communication. In this thesis, the main contribution is focused on the characterization and measurement of communication channels in the mmWave band. To this end, it proposes the study of the physical layer of the communication channel from different perspectives, such as optimization, simulation, emulation, classification, characterization and measurement. Four main contributions can therefore be identified. First, the use of optimization algorithms based on evolutionary algorithms has been proposed to optimize key performance indicators in a simulated network deployment. The simulation of propagation channels through simulators and channel modeling simplifies the complexity of the analysis compared to costly measurement campaigns, making it a suitable solution in the early stages of network development. Optimizing network parameters from a simulator thus provides a first approximation to the requirements of a real communication network. By optimizing the transmit power in the radiating elements of the network, it is possible to simultaneously improve several network parameters at multiple layers of such network. Second, channel emulation and identification techniques have been developed for the emulation and classification of propagation channels. On the one hand, channel emulation, based on the reconstruction of target channels in the laboratory, is an intermediate solution between theoretical simulators and complex measurement campaigns. The recreation of scenarios with specific characteristics is useful for testing devices under specific operating conditions and for analyzing the communication channel. This thesis contemplates the time-gating technique as a postprocessing method for channel emulation. On the other hand, scenario identification is crucial to understanding the environments in which communications take place and deciding how they should be conducted. In this thesis, dimensionality reduction techniques are considered as a classification strategy for several propagation scenarios. Both channel emulation and scenario classification techniques have been validated through measurements in controlled environments. Third, the knowledge of the channel response with respect to an incident wave is essential for establishing communication on a link. Therefore, it is necessary to characterize the spatial and temporal profiles of the communication channel by determining the direction-of-arrival and time-of-arrival of the channel multipath components. For this reason, this thesis develops techniques based on the joint estimation of the direction-of-arrival and time-of-arrival based on frequency invariant beamformers, which are robust to the large bandwidths employed in the mmWave band. In particular, two different techniques have been proposed, one based on elliptical arrays for 2D characterization, and the other based on toric arrays for 3D characterization. Finally, the thesis concludes with the presentation of a measurement campaign in an industrial environment under operational conditions. The characterization of the propagation channel based on simulations and channel emulation is complemented and validated by measurement campaigns in the final stages of the development of a communication network. These industrial environments have become particularly relevant due to industrial processes automation in factories. However, these environments are particularly challenging due to their time-varying and high-density nature, leading to the high probability of link blockage. Therefore, this thesis examines the radio propagation conditions in such environments and analyses how network management techniques provide coverage in these challenging scenarios.