Jiann-Ching Guey

On 5G radio access architecture and technology [Industry Perspectives]

In early 2012, the International Telecommunication Union (ITU) under the United Nations initiated a program to develop an International Mobile Telecommunication (IMT) system for 2020 and beyond (IMT-2020), thereby officially kicking off the global race toward a yet to be defined fifth generation (5G) mobile network. Fast forward three years, the vision of this next-generation system is beginning to take shape. A recent ITU-Radiocommunication Standardization Sector (ITU-R) Draft Recommendation [1] on IMT-2020 identifies three key usage scenarios in 5G as enhanced mobile broadband, massive machine-type communications, and ultra-reliable and low-latency communications. The same report also issues guidance on the requirements for these scenarios in terms of performance index such as spectrum efficiency, latency, connection density, and area traffic capacity, as shown in Fig. 1. In order to meet these requirements and bring this visionary system to reality, we believe that the future 5G network will be one that is built on a small cell backbone. As spectrum suitable for mobile communication becomes more and more scarce, densification is the only way to meet the area traffic capacity demand. Even for millimeter-wave band, where spectrum is abundant, the channels propagation characteristics will likely limit its range for mobile access to that of a small cell, at least in the early phase of 5G before device technology matures. The small cell also brings the radio access point closer to the end device, thereby shortening the last and most challenging segment of an end-to-end communication link, consequently reducing latency and increasing reliability. Many of the massive number of machine-type communications can also benefit from the extended battery life resulting from shorter uplink distance. To facilitate the deployment of a future mobile network that has the small cell as its primary traffic bearing workhorse,the current radio access architecture needs to undergo some major revamping, and new technologies need to be introduced.

Beam Space Selection for High Rank Millimeter Wave Communication

We address beam space reduction method inside which proper analog beamformers are selected for hybrid beamforming millimeter Wave (mmWave) system. We apply insights from spatial multiplexing theory on beam subspace selection and show that computational effort of searching for desired beamformers can be significantly reduced. A simplified channel throughput estimation measure is also devised to further lessen the complexity in the course of beamformer searching. Simulation results demonstrate that near-optimal performance can be achieved with significantly reduced complexity.

A two-user approximation-based transmit beamforming for physical-layer multicasting in mobile cellular downlink systems

Multicast has been known as an efficient transmission technique for group-oriented applications such as multi-party video conferencing, video streaming for paid users, online gaming, and social networking. In this paper, we investigate physical-layer multicasting in mobile cellular downlink systems, where the antennas at base station are employed to transmit common signals to multiple users simultaneously. A central design problem of downlink physical-layer multicasting is the search for the optimal beamforming vector that maximizes the multicast rate. Traditionally, the problem has been formulated as a quadratically constrained quadratic programming problem and shown to be NP-hard in general. In this paper, starting from examining the Karush–Kuhn–Tucker stationary conditions, a new method based on two-user approximation is proposed for the search for the optimal beamforming vector. The method is able to achieve a much higher multicast rate than the existing methods and provides an attractive trade-off between performance and complexity, especially for the case of using a large number of antennas. Using a large number of antennas at base station, also known as the large-scale multiple-input and multiple-output technique, has been regarded widely as one of the most promising technologies to increase system capacity, coverage, and user throughput for future generations of mobile cellular systems.

Modeling and evaluation of beam tracking in mobile millimeter wave communication

The most critical challenge facing mobile millimeter wave communication is the precise projection of high-gain and narrowly focused radiation toward the multi-paths leading to a device as it moves and rotates in the field. In order to design and evaluate such beam tracking mechanism, an accurate model of the propagation environment is required. Unfortunately, conventional channel models for system operating in carrier frequency lower than 6 GHz are typically statistically based with time-invariant parameters and thus inadequate to simulate the dynamics in spatial characteristics the millimeter wave channel may exhibit. In this paper, such shortcoming is overcome by the introduction of certain time-varying parameters to a widely adopted geometry based stochastic model. With the revised model, we proceed to tune its parameters to fit the statistics extracted from the latest measurement data and subsequently investigate the models capability in evaluating beam tracking performance. The beam tracking algorithms are generic in nature and only assume certain imperfect knowledge of the channels spatial profile. It is found that at 28 GHz carrier frequency, beam tracking can be maintained at a moving speed of 60 kmph and an angular speed of 720°/sec with around 20% loss comparing with the stationary reference if the spatial profile is updated at a rate of 50 Hz.

A New Transmit Beamforming Technique for Physical-Layer Multicasting in Cellular Downlink Systems

This paper investigates the design of physical-layer multicasting in the cellular downlink systems, where antennas at a base station are used to transmit a common signal to multiple single-antenna users. The central design problem of the physical multicasting is the search of the optimal beamforming vector that maximizes the multicast rate and has been known as an NP-hard problem in the literature. In this paper, a new method based on a two-user approximation is proposed for optimizing the beamforming vector. The method is shown to achieve a much higher multicast rate than the existing methods and provides an attractive tradeoff between performance and complexity, especially for the case of using a large number of antennas.