The hype machine is in full swing with the latest standard in mobile telephony — 5G. This is the fifth generation of wireless, mobile broadband communications. It should not be confused with 5 GHz Wi-Fi, which is sometimes erroneously referenced as 5G. This 5G is a whole new ballgame and promises to be a revolution in how we live our daily lives — not merely an evolution of versions past.
The cellular communications world self-references in terms of which generation of the technology is being used for signaling. The first generation, 1G, brought us voice in a mobile platform; 2G brought us text messages, 3G brought us the mobile Internet, and 4G long-term evolution (LTE) gave us a speed increase of 10x over 3G.
Although, the theoretical maximum data rates delivered with 4G modems top out at a remarkable 2 Gbps, 4G download speeds here in the U.S. average only about 30 Mbps, as shown in Figure 1 — and upload speeds are significantly lower. The upstream side is generally used as a control plane rather than a data plane.
To put this in perspective, a 1080P HD movie requires about 5 Mbps to stream without significant buffering delays. With aggressive compression, Netflix recommends at least 25 Mbps for 4K format video (3840 × 2160 pixels). This means that 4K video is not achievable with the average mobile broadband speeds here in the U.S. even though many modern smartphones are capable of 4K screen resolution.
5G NR (new radio) modems have a maximum data rate of 20x the maximum 4G LTE speeds, so 5G could theoretically deliver up to 20 Gbps wirelessly. More realistically, we could expect speeds in excess of 200 Mbps in the short run and much faster as the network gets built out, as shown in Figure 2. This can potentially open new markets such as 4K streaming to mobile devices like smartphones and tablets.
The cellular carriers (“telcos” or telephone companies), to increase market share, emphasize the speed and reliability of their networks as their primary selling points. Since the smartphone market is fairly saturated at this point, the only way for the carriers to increase revenue is to either steal subscribers from other carriers, typically via cost cutting, or to expand into new market areas. Of all of the potential new markets, eliminating the “last mile” problem and servicing the growing Industrial Internet of Things (IIoT) market (estimated to be in excess of $751.3 billion by 2023) are both hot possibilities.
In order to provide service to an individual’s home or business (the last mile), the service provider needs to get the signal out of their network and deliver it to the destination. For the telcos this means from the local exchange’s facility to the destination (referred to as the point of demarcation or the DMARC). The DMARC is where the service provider’s network ends and the customer premises equipment (CPE) begins.
This is troublesome because this last mile is traditionally serviced with copper or fiber-optic cabling, which is incredibly expensive to lay and to maintain. Therefore, it would behoove the service provider to deliver high-speed service wirelessly and eliminate the cost of the fixed cable plant. This also enables delivery of services to areas where it would otherwise not be cost effective due to low customer density, such as in rural areas or to new housing or building developments. In the telco business, this wireless model is referred to as fixed wireless access (FWA) because the destination is fixed, not mobile, as shown in Figure 3. The FWA market is expected to grow at a Compound Annual Growth Rate (CAGR) of approximately 84% between 2019 and 2025 eventually reaching an estimated $40 Billion.
Once service can be established, the cellular signal becomes a means to deliver additional services known as over the top (OTT) services. For example, to be the carrier offering Netflix, Hulu, or YouTube TV services to customers rather than just traditional television channels. This gives the customer additional choices with a lower price than what they would find with many of the classic bundling services found in the cable industry. We could even see the cable industry using the telco’s infrastructure to deliver OTT services to traditional cable service customers.
With FWA, service into the customer’s premises could be accomplished via a 5G to Wi-Fi gateway. Another option is bringing the 5G service directly into the premises via a signal booster or repeater. However, that would supplant existing Wi-Fi interfaces in devices such as laptops and other consumer electronics and would require a significant change in hardware design.
The Real Jackpot
While the carriers like to throw around numbers related to how fast the 5G network will be for users’ mobile devices, their real bonanza is in machine-to-machine (M2M) communications for the Internet of Things (IoT). This market segment is expected to experience a CAGR of 54.3% from 2021 to 2025, as shown in Figure 4. However, IoT devices do not need multi-megabit services. Many IoT applications are fine with data rates in the tens of kilobits per second. However, IoT applications used to control devices like door locks, interior lighting, industrial process controls, and robotics are latency sensitive.
Latency is the amount of time between when a command is sent and when data actually starts to flow. Users experience this as a lag between when they hold their access card to the door reader and when the door actually unlocks for them. A second of lag might be acceptable, but 20 seconds is not tolerable. Each M2M application has its own acceptable latency, ranging from a few milliseconds to even microseconds. Typical 4G LTE networks, have latencies from 50 to 100 ms. While that is likely acceptable for turning on overhead lights, it would be intolerable for applications like autonomous vehicles. For example, a vehicle traveling at 60 mph (96.56 km/h) would traverse almost 8.8 ft (2.68 m) in 100 ms. This could be disastrous if the vehicle was trying to determine if the object ahead was a pedestrian and it needed to stop quickly. So, autonomous vehicles are another of the primary markets for 5G technology with its 1 ms latency.
There is another dimension to consider for industrial applications — reliability. The current 4G network is at 99.8%. So, a dropped call rate of 0.2% is considered a super-reliable, carrier-grade mobile network. But, the reliability requirement for industrial applications is typically estimated at six-9s (99.9999%). 5G networks can deliver that level of reliability by using cell duplication, the clever use of radio spectrum, and massive MIMO (multiple-in, multiple-out) antennas.
5G and the Radio Spectrum
The Third Generation Partnership Project (3GPP) is an international standards organization that develops protocols for mobile telephony. Their Release 15 focused on 5G for consumers and is the current standard. Release 16 shifts the focus from the consumer over to the IIoT and other industrial markets. This release is expected sometime in late Q1 or early Q2 of 2020. In the meantime, the carriers are focusing on network build-out and preparing for the transition from 4G to 5G.
The 5G NR utilizes a variety of different radio frequency bands and can leverage 4G LTE as a control channel. The 3GPP refers to the 5G frequencies as Frequency Ranges 1 and 2 (FR1 and FR2). FR1 frequencies range from 410 MHz to 7.125 GHz, and FR2 from 24.250 to 52.600 GHz (mmWave). However, common carrier parlance refers to 5G frequencies as low, mid, and high band.
5G radio access technology (RAT) currently targets two different deployment modes to take advantage of existing 4G LTE infrastructure. The first is non-stand-alone mode (NSA), which leverages the existing 4G network for the 5G control plane. Essentially, the 4G network would be used to set up and tear down calls (the control plane) while the 5G network would provide data transfer (the data plane).
In standalone (SA) mode, 5G cells are used for both signaling and data transfer. This would include the new 5G Packet Core architecture instead of continuing to rely on the existing LTE network. The 5G Packet Core is expected to be more efficient, have lower cost, and support the development of new use cases.
Since 5G allows the service provider to use multiple frequencies from their licensed spectrum, the low (sub 1-GHz) and mid (1 GHz – 7.125 GHz) frequencies will be used for extended range and rural coverage where those frequencies provide for greater distance between cellular towers and provide better penetration into foliage, buildings, and other radio-blocking obstacles. Even though there is not huge available bandwidth in these frequencies, the 5G NR waveforms will still support greater data rates than 4G in these same frequencies. Real-world tests using 5G enabled phones show that low-band 5G (600 MHz) is typically 20 – 50% faster than 4G LTE, with occasional locations showing in excess of 100 Mbps.
Much of the hype related to speed of 5G is centered on the FR2 frequency range. There is significant available bandwidth in these mmWave frequencies and tests on Verizon’s 5G mmWave network show downstream data rates in excess of 1 Gbps.
However, mmWave signals are very sensitive to range from the cell tower. Standing within a few hundred feet or so of the tower can result in speeds over 1.5 Gbps as long as you are facing the tower and not moving. However, they are easily blocked by your body, foliage, moving automobiles, heavy rain, buildings, and even windows. In a less than optimal location, the data rates drop significantly to just around 200 Mbps. That is still very fast in comparison to 4G LTE, but nowhere near the optimal case.
Fortunately, the size of a mmWave cellular transceiver is considerably smaller than the typical 4G macrocell and can be easily mounted on streetlight-sized poles, as shown in Figure 5. That being said, you will need a cell site roughly every 1–2 blocks to be able to keep from switching back to 4G. This will be a significant build-out for the carriers. And, since different carriers have licensed different portions of the mmWave spectrum, you might end up being locked into a particular carrier with a particular phone to enjoy the benefits of 5G service to your mobile handset.
5G is not just one thing. Just what 5G actually is depends greatly on the carrier and where you are located. Although 5G services started to become available in select areas of select cities during 2019, the network is not expected to see significant nationwide build-out for mmWave services until 2021-2022. You can get low and mid-band 5G services in the meantime, which will be faster than the 4G service, but the speed increase will not be leaps and bounds faster on the lower bands.
5G promises to bring a whole new set of services to the consumer: virtual reality, low-latency gaming, streaming 4K video, and more. But consumer applications do not represent the significant payback for investment in the expansion of the network.
The new markets for 5G will be the FWA and low-latency IIoT applications. The IIoT alone is expected to more than double carriers’ revenues from 5G services over the next 10 years. And, as new narrow-band IoT (NB-IoT) 5G services are rolled out, there could be billions of low-end sensors such as parking meters contributing additional revenue.
Does 5G cellular represent a connectivity revolution? If it lives up to the promises of the 3GPP, it certainly might. But the devil will be in the details and whether 5G is able to penetrate the market segments that the industry is predicting.