by Terry Edwards, Engalco Research
The now burgeoning 5G communications networks all involve much higher information rates than the earlier 4G LTE. This situation leads to Gbps rate requirements for the “xhaul” (backhaul and fronthaul) site-to-site connections and the basic choice is between fiber-optic or millimeter-wave (mmWave) transmission technology. The approach chosen for any specific location is determined by cost, environmental considerations and speed of installation.
In many urban and suburban areas, networks utilizing mmWaves have been implemented and this approach is growing strongly (e.g. Verizon in the USA).
mmWave radio links can be installed comparatively rapidly and in a less costly manner than most optical fiber cabling. Available bandwidths (upwards of several hundred MHz) are much smaller than those applying to fiber, but these bandwidths are eminently suitable for 5G xhaul purposes. Hence mmWave technology in this role has for several years been prominent and this situation is expected to expand for many years to come.
RF signal transmission up to and through microwave frequencies (“topping-out” around 18 GHz) suffers relatively little in terms of:
- Attenuation due to atmospheric effects
- Penetrating walls and other similar obstacles
However, mmWave transmission certainly does suffer in terms of both aspects 1. and 2. cited above.
An examination of atmospheric attenuation (dB/km) as a function of frequency, from 1 to 100 GHz, reveals highly non-linear behavior due to specific oxygen and water-vapor resonances around certain fairly narrow frequency bands. The “standard graph” is shown as Figure 1.
Through microwave frequency bands, up to about 18 GHz, the atmospheric attenuation is relatively small and increases only slowly with frequency—reaching approximately 0.02 dB/km @ 18 GHz. There is a significant water vapor resonance (increased RF power absorption) @ 23 GHz and this explains the accelerating rise in attenuation reaching a maximum of 0.35 dB/km @ 23 GHz. Between about 26 and 45 GHz, the attenuation is relatively flat—remaining below 0.2 dB/km. This region includes “Ka-band” which is also important for fixed wireless access (FWA).
Above 45 GHz, the attenuation begins to rise steeply, reaching a maximum 12 dB/km at 60 GHz. This is the first peak due to oxygen molecular absorption.
In spite of this substantial attenuation, the 60 GHz band (V-band) is a very popular option because it is unlicensed.
Further increases in frequency above 60 GHz indicate rapidly falling attenuation until the next “quasi-plateau” is encountered, ranging from 71 to 86 GHz. This range includes the lower and upper E-bands (as well as the 77 GHz vehicle radar band).
The W-band (94 to 95 GHz) is also important and the top frequency in this band (i.e., 95 GHz) is well below the next rapid increase characteristic. Through the W-band the attenuation maximizes at around 1 dB/km.
To a considerable extent, these characteristics determine choices of mmWave frequency bands. There are, however, further important characteristics and features.
On top of atmospheric attenuation, precipitation causes further losses on signal strength (i.e., further attenuation). Rain and snow must be allowed for when designing any mmWave link—losses of several additional dB will regularly be an important consideration in the “link budget” (RF power levels across any transmission link).
Another challenging feature with mmWave transmission is the fact such signals cannot penetrate solid walls. With brick or concrete walls the attenuation approaches total blocking. Even less substantial internal walls result in considerable additional path loss between the transmitter and receiver. Senic et al  reported measured and simulated data for path loss across an indoor link operating at 83.5 GHz (upper E-band). The basic results are summarized in Figure 2.
For a 20 ft. link, the median path loss was 87 dB. When the link is extended ten-fold i.e. to 200 ft., the loss is massively increased to approximately 104 dB. It is important to appreciate these data are median. Substantial scatter was observed  around all loss data, ranging from around 4 dB @ 50 ft. distance to as high as about 13 dB @ 500 ft. distance.
As a result of the above challenges, mmWave links are almost always much shorter than microwave links – typically around 1,000 ft. maximum outdoors or around 200 ft. maximum indoors. This means networks utilizing mmWave links are necessarily provided with appropriate high-gain antennas and associated hardware. Due to the relatively short wavelengths, the terminal equipment is almost always physically much smaller than that required for microwave links.
In common with all communications links, mmWave-based systems require transmit-receive modules. Each module contains a mmWave chipset (together with digital circuitry) connecting electrically to the antenna structure. A photograph of a typical mmWave transceiver is shown in Figure 3.
The module is packaged within the metal enclosure to the left in Figure 3. Maja Systems offers the electronics of this product in the form of a single CMOS chip (the MW6022) and this chip includes integrated ADC and DAC sections as well as a modem. Data rates up to 4.25 Gbps are supported and the DC power consumption is less than 350 mW. To replace the traditional horn antenna (top right in Figure 3) Maja Systems now offers a unique miniat,urized form of antenna structure.
Companies manufacturing mmWave transceiver modules (covering various frequency bands) include, for example:
- Axxcss Wireless Solutions
- BridgeWave Communications
- E-Band Communications
- Huawei Technologies Co.
- Maja Systems
- Nokia Corporation
- Wireless Excellence
Operating center frequencies of the products available from these OEMs vary. Ka-band (mainly 28 GHz) and V-band (60 GHz) are the most commonly encountered. A recent industry report provides substantial information on mmWave transceiver module markets . This report reflects the first half 2020 global industry downturn resulting from the economic impact of Covid-19. A substantial second half rebound was assumed, which is in line with most current expectations.
1. Jenena Senic, Camillo Gentile, Peter B. Papazian, Kate A. Remley and Jae-Kark Choi, “Analysis of E-band Path Loss and Propagation Mechanisms in the Indoor Environment,” IEEE Trans. Antennas Propag., 2017, v 65, pp 6562 – 6573.
2. mmWave Transceivers into 5G ‘xhaul’ – Regional & Global Markets: 2019 – 2028. Engalco Research, July 2020.