Millimetre wave frequency band as a candidate spectrum for 5G network architecture: A survey

Abstract

In order to meet the huge growth in global mobile data traffic in 2020 and beyond, the development of the 5th Generation (5G) system is required as the current 4G system is expected to fall short of the provision needed for such growth. 5G is anticipated to use a higher carrier frequency in the millimetre wave (mm-wave) band, within the 20 to 90 GHz, due to the availability of a vast amount of unexploited bandwidth. It is a revolutionary step to use these bands because of their different propagation characteristics, severe atmospheric attenuation, and hardware constraints. In this paper, we carry out a survey of 5G research contributions and proposed design architectures based on mm-wave communications. We present and discuss the use of mm-wave as indoor and outdoor mobile access, as a wireless backhaul solution, and as a key enabler for higher order sectorisation. Wireless standards such as IEE802.11ad, which are operating in mm-wave band have been presented. These standards have been designed for short range, ultra high data throughput systems in the 60 GHz band. Furthermore, this survey provides new insights regarding relevant and open issues in adopting mm-wave for 5G networks. This includes increased handoff rate and interference in Ultra-Dense Network (UDN), waveform consideration with higher spectral efficiency, and supporting spatial multiplexing in mm-wave line of site. This survey also introduces a distributed base station architecture in mm-wave as an approach to address increased handoff rate in UDN, and to provide an alternative way for network densification in a time and cost effective manner.

Introduction

The vast proliferation and the enormous use of data-hungry devices such as smart-phones and laptops have dramatically increased Global Mobile Data Traffic (GMDT). GMDT already grew up to 74% in 2015, reaching 3.7 exabytes per month. This figure is forecasted to grow by 2020 to 30.6 exabytes [1]. The fourth generation (4G) system was not designed to cope with this huge growth. Therefore, the next generation of wireless standards needs to extend its bandwidth into a higher carrier frequency in the millimetre wave (mm-wave) band ranging from 3 to 300 GHz.

The spectrum at microwave bands (MW) is becoming tooscarce, because almost all cellular communication systems are operating in the sub 3 GHz band. Therefore, moving to the mm-wave bands is essential because there are wide unused bandwidths, particularly from 20 to 90 GHz. These bands could become accessible for the Fifth Generation (5G) system as a potential solution for achieving a 1000 folds capacity increase compared to the current Long Term Evolution Advance (LTE-A) networks [2], [3]. The mm-wave spectrum refers to the frequencies from 30 to 300 GHz, which is called the extremely high frequency (EHF) band. The 3–30 GHz spectrum is called the Super High Frequency (SHF) centimetre wave band. Because EHF and SHF bands have approximately similar propagation conditions, the 3–300 GHz spectrum is collectively called the mm-wave band with their wavelengths ranging from 1 to 100 mm [4]. The high speed data rate and low end-to-end latency requirements cannot be fulfilled with mere evolution from the existing 4G network or minor changes [5]. Therefore, researchers focus their attention on technologies that comprise major and radical changes in the base stations (nodes) level as well as at the network (core, backhaul) level, because only these types of changes have the capacity to meet these stringent requirements. In this context, the mm-wave band is the most prominent technology and the key enabler to satisfy the extreme demands of future applications.

In this paper, we have carried out a detailed survey regarding the use of the mm-wave band for cellular purposes by considering most of the recent research contributions and publications in this field. The focus is on the impact of using the mm-wave band in a new network architecture, which requires a radical change at both network and component levels. New network design options have been proposed in many contributions in order to cope with the propagation characteristics of mm-wave. These include: higher order sectorisation, ultra-dense network implementation, and the use of distributed base station architectures. These architectures exploit the mm-wave to significantly boost the network capacity, and to cope with severe losses that characterise mm-wave propagation.

After presenting the potential bandwidth allocations in Section 2, we have summarised the characteristics of mm-wave communications in Section 3, which include high path loss and atmospheric attenuation. Due to their short wavelengths, mm-wave signals have severe propagation loss and scattering, and high sensitivity to blockage by obstacles such as buildings, street furniture, and human bodies. We then introduce the rain effect, as mm-wave signals can suffer significant attenuations in heavy rain as raindrops have approximately the same size as the signal wavelengths and thus result in severe signal attenuation due to scattering. One of the solutions to this loss is by reducing the Inter-Site-Distance (ISD), which will decrease the signal path by making the Access Point (AP) much closer to the users. After that, three wireless standards operating in the mm-wave band are introduced in Section 4. These standards are operating in the 60 GHz band and should provide very high data speed in short ranges. In Section 5, we have identified the potential use of mm-wave in 5G, for indoor, outdoor, and backhaul solutions, and the contributions in each field are highlighted and discussed. Relevant issues to adopting mm-wave in 5G are discussed in Section 4, which include the handoff issue and interference in Ultra-Dense Network (UDN), waveform consideration, providing spatial multiplexing in Line of Site (LoS) transmission, supporting Machine-to-Machine (M2M) traffic, and introducing distributed base stations in the mm-wave environment. Finally, the conclusions are drawn in Section 6.