Delivering 5G mmWave fixed wireless access

& -September 28, 2017

Mobile network operators view fixed wireless access (FWA) as the first application for 5G in millimeter wave (mmWave) spectrum. FWA is poised for rapid roll-out, but there remains a challenge: building power-efficient and cost-effective base stations and customer premises equipment capable of supporting the high power and bandwidth.

FWA is a compelling first application for several reasons. Operators can deliver multi-gigabit speeds to homes, apartments, or businesses in a fraction of the time and cost of traditional cable and fiber-to-the-home installations. This opens new markets that previously were available only to satellite, fiber, or cable providers. An additional attraction is the potential to directly extend FWA infrastructure investment towards providing 5G mobile access.

Not surprisingly, Verizon, AT&T, and other carriers are aggressively trialing FWA, with the goal of full commercialization in 2019. Market-research firm SNS Telecom predicts that FWA will take off rapidly, generating $1 billion during 2019 alone. [1]

FWA basics
Fixed wireless access describes a wireless 5G connection between a centralized base-station and numerous fixed or nomadic user locations (Figure 1). FWA is considered particularly suitable for several environments, including:
  • Dense urban and suburban areas: With the high capacity and throughput provided by FWA, operators see the potential to deliver high performance to large numbers of users while generating sizable revenue.
  • Indoor/outdoor hotspots: FWA can provide high-speed, low-latency connectivity to support crowds. For example, sports fans in stadiums could experience immersive live-stream video taken from every angle while watching the game.

Figure 1 End–to–end fixed wireless access networking using MIMO and beamforming

FWA spectrum, bandwidth, and regulatory progress
Operators have already taken steps to meet their first FWA challenge: obtaining the necessary spectrum. Most deployments are expected to use mmWave frequencies, where large swaths of contiguous unpaired bandwidth are available at very low cost. For example, Verizon spent approximately $0.017 per MHz-PoP for Straight Path Communication’s 28 GHz and nationwide 39 GHz spectrum [2]. That was a bargain compared to the AWS-3 paired spectrum that averaged $2.72 per MHz-PoP [3] in 2015.

Global mmWave spectrum availability is shown in Figure 2. The first FWA deployments are expected to use the 24.25–29.5 GHz and 37–43 GHz range.

Figure 2 mmWave and cmWave Global Bands

The 3GPP 5G New Radio standard (Release 15) has proposed component carrier bandwidths for mmWave (a.k.a. frequency range 2) to be up to 400 MHz, and early commercial base stations may support over 1.2 GHz of instantaneous bandwidth using carrier aggregation [4].

In the U.S., the plan is to make use of about 3850 MHz of mmWave spectrum for 5G—roughly six times the entire amount of bandwidth previously allocated for commercial use. The Federal Communications Commission (FCC) has already defined 28 GHz and 39 GHz bands for 5G [5] and is exploring additional spectrum between 4 GHz and 24 GHz [6]. The FCC has also set very high transmit power levels for base stations and mobile devices (Table 1).

Table 1 FCC power limits for 28GHz and 39GHz bands

With regulatory rules defined, large amounts of spectrum already in the operators’ hands, and enough power output to deliver Gbps over several city blocks, the industry is ready to start building base stations.

FWA base station design
FWA base stations present several RF design challenges. To deliver the needed capacity across the sectored cell coverage with sufficient indoor penetration, base stations must utilize sophisticated active phased array antenna systems (AAS). It is essential that the active antenna array is highly power-efficient and robust to allow passive cooling of all-outdoor tower top electronics. Furthermore, base stations must be lightweight and compact enough to mount in locations ranging from traditional cell towers to streetlamp poles.

The size, or number of active channels, in the phased array is a critical consideration. Array size depends on the scanning (azimuth and elevation) requirements and the desired effective isotropic radiated power (EIRP). EIRP is the product of the number of active channels, the conducted transmit power of each channel, the beamforming gain (array factor), and the intrinsic antenna element gain.

The required scanning range depends on the deployment scenario (Figure 3). For suburban deployments, a fixed or limited scan range (< 20°) in the elevation plane may be sufficient, making it possible to take advantage of higher intrinsic antenna element gain. For “urban city” deployments, wide-scan ranges are needed in both azimuth (~120°) and elevation (~90°).

Because of the limited scan range, a suburban deployment can use antennas with 6dB higher intrinsic element gain than an urban deployment. As a result, a suburban phased array requires half as many active channels to achieve the same EIRP, significantly reducing power and cost. Conversely, for the same power and number of active channels, a suburban array can achieve 4× higher EIRP.

Figure 3 Array complexity depends on the scanning range needed for the deployment scenario

At mmWave frequencies, the lattice spacing between phased array elements becomes extremely small – at 40GHz, it is less than 4mm. To minimize feed losses, it’s paramount to locate critical front-end components as close as possible to the radiating elements. This is driving a push to shrink the power amplifier (PA) footprint and integrate as many functions as possible into single- or multi-chip modules.

In addition, recent trends toward integrated dual polarity and transmit/receive are driving a further 4× increase in circuit density (Figure 4). Early mmWave base station designs often used separate single polarization transmit and receive antenna arrays, which allowed twice as much area for components and avoided the additional insertion loss and linearity challenges of a single pole double throw (SPDT) transmit switch.

However, adaptive beamforming is critically dependent on the ability to calibrate the receive and transmit arrays relative to one another. This calibration depends on the spatial correlation between the transmit and receive arrays. It is therefore preferred to integrate transmit/receive so that the array shares a common set of antenna elements and RF paths. Additionally, dual polarization is needed for diversity and capacity.

Figure 4 The evolution of the FWA antenna array

Tiling all these functions in such a small area requires either very small PAs based on SiGe/Si semiconductor technology, which produce very low power output, or the use of high power density semiconductor technologies like GaN, which has more than 100 times the power density of SiGe/Si [7].

Loading comments...

Write a Comment

To comment please Log In