Page 13 - EngineerIT February 2022 UPDATED
P. 13

ICT – MILLIMETRE WAVE TECHNOLOGY


        Why millimetre wave requires a




        different approach to DPD and how




        to quantify its value




        By Hossein Yektaii, Wireless System Architect; Patrick Pratt, Algorithm Design Engineer and Frank Kearney, Engineering Manager




                                             Introduction
         In this article
         In the 5G New Radio standard,       In addition to reduced latency and higher reliability, the exponential increase in demand
         millimetre wave (mmWave)            for higher data throughput has been one of the most powerful driving forces behind 3GPP
         frequencies, in addition to sub-6   5G NR standard.
         GHz frequencies, are utilised to      While the 4G LTE systems were deployed in sub-3 GHz bands, the allocation of the
         enhance throughput. The use of      new spectrum in the 3 GHz to 5 GHz range in recent years has enabled wider channel
         mmWave frequencies provides         bandwidths (BW) in 5G NR. Compared to the 4G LTE, the maximum channel bandwidth
         unique opportunities for a drastic   has increased from 20 MHz to 100 MHz in sub-6 GHz frequencies. Besides wider channel
         increase in data throughput while   bandwidth, multiple transmit and receive antennas and, ultimately, massive MIMO
         presenting new implementation       technology, has further increased the spectral efficiency. While all these improvements
         challenges. This article explores   help to deliver higher data throughputs, the fundamental limitation—the relatively small
         architectural differences between   amount of allocated sub-6 GHz spectrum—continues to limit the peak throughput for
         sub-6 GHz and mmWave base station   individual users to less than 1 Gbps.
         radios, with particular emphasis      In 5G NR, for the first time in the history of the 3GPP standards, millimetre wave
         on the challenges and benefits of   frequencies between 24,25 GHz to 52,6 GHz are allocated for cellular mobile applications.
         implementing digital predistortion   This new frequency range is referred to as FR2, in contrast to the sub-6 GHz frequencies
         (DPD) on these systems. While       termed FR1. There are much larger swathes of spectrum available in FR2 relative to FR1. A
         digital is a well established       single channel in FR2 could be as large as 400 MHz, enabling unprecedented throughput.
         technique commonly used in sub-     However, the use of mmWave frequencies brings new implementation challenges to both
         6 GHz wireless communication        the base station (BS) and user equipment (UE). The most significant of these challenges
         systems to improve the power        are higher path losses and lower PA output powers, making the link budget between the
         efficacy, most mmWave radios do     base station and UE quite challenging.
         not use DPD. Using a prototype 256    Path loss between BS and UE is defined as P l [dB] = 10log 10 (P t/P r), where P t and
         element mmWave array, built with    P r are the transmitted and received power, respectively. In free space, the received
         ADI beam formers and transceivers,   power is a function of distance and wavelength, also known as Friis’ formula, where P r
         we are able to demonstrate that     (d,λ) = P t G t G r (λ/4πd)², and G t and G r are the transmitter and receiver antenna gains,
         DPD improves the effective isotropic   respectively. λ is the wavelength and d is the distance between transmitter and receiver.
         radiated power (EIRP) by up to 3    In a typical wireless communication environment, due to reflection off nearby objects
         dB. This allows for a 30% reduction   and loss through construction material, the path loss is much more complex to model and
         in the number of array elements,    estimate. However, to form an understanding of higher path loss at mmWave frequencies
         relative to an array without DPD, for   compared to sub-6 GHz, let us assume free space propagation, similar antenna gains, and
         the same target EIRP.               equal distances between base station and user equipment. Using this approach, the path
            The purpose of this article is   loss at 28 GHz compared to 900 MHz is 10xlog (28000/900)² = 29.8 dB higher!
         to draw a comparison between a        It is not uncommon for base station power amplifiers at sub-6 GHz frequencies to
         traditional sub-6 GHz macro-cellular   output tens of watts of RF power with efficiencies above 40%. This is enabled by the
         and a mmWave base station radio     adoption of high efficiency PA architectures such as Doherty and the use of advanced
         and antenna design. It further covers   digital predistortion techniques. In contrast, the highly linear class AB mmWave PAs
         how these design differences impact   typically output less than 1 W of RF power and have single digit efficiencies. At mmWave
         DPD implementation in mmWave        frequencies, these operating conditions exacerbate the link budget challenges between
         arrays relative to sub-6 GHz radios.  BS and UE. The solution to both challenges—the larger path loss and lower power per
                                             PA—is the more accurate delivery of power to specific spatial locations. This is achieved



                                                  EngineerIT | February 2022 | 11
   8   9   10   11   12   13   14   15   16   17   18