Hi. My name is Harald Remmert. I'm a Director of Engineering at Digi International responsible for R&D of Cellular & Networking Products. Digi International is the leading global provider of mission-critical, machine-to-machine, and Internet-of-Things connectivity products and services. In today's webinar, we will explore how you can leverage LTE for your applications today and prepare for 5G in the future.
I need connectivity. Remember the good old 2G days? Only a single option for cellular connectivity, four bands for worldwide coverage. Life was simple. And Sam Elliott was in charge. Fast forward to today, the world is migrating from 2G and 3G cellular technology to 4G and on towards 5G. The number of cellular bands is exploding. From four bands for worldwide coverage to over 70 bands and counting. From technology optimized to transmit a few bytes a day to a gigabit, to LTE speeds surpassing broadband connectivity. It feels like herding cats. And Sam Elliott is nowhere to be found. So it really changed from I need connectivity to what connectivity do I need.
Today, cellular network support 2G, 3G, and 4G devices. 2G is a 30-year-old technology. 3G is already 20 years old. 4G is here to stay and will evolve to 5G in the next few years. 5G offers new use cases and is further diversifying the cellular landscape. From a significant increase of the number of devices, to battery powered devices with 10-year battery life, to high power multi-radio devices, from technology optimized to transfer a few bytes a day to a Gigabit LTE speeds surpassing broadband connectivity, to latency with latencies as low as one millisecond infrastructure latency for real-time industrial control to multi-second latency for extra coverage and on and on installation. So 5G is really exciting but let's talk about the 2G and 3G technology sunset first.
In many cases, the migration away from 2G and 3G is driven by the desire to re-purpose spectrum for more efficient LTE data traffic. It's more cost-effective for a carrier to operate an LTE network than a 2G or 3G network. Prior to sunsetting a network, carriers may regionally re-purpose spectrum or tune the radio access network. In these cases, your 2G or 3G device may lose connectivity.
Timelines vary by country or region. Here's an example of the 2G and 3G sunset in the U.S. Sunsets have happened or are imminent. Transition from 2G and 3G to 4G LTE now. In Europe, 2G networks are projected to outlive 3G until 2025. However, carriers may offer incentives to migrate sooner. In Asia, 2G and 3G sunsets have happened or are imminent. For example, in China, Singapore, Australia, these have already happened. Telstra has shut down their 2G network in 2016, and then 3G is projected to be shut down in 2020.
Now, with these in mind, let's explore how we can move forward and how to pick the right technology. So, in order to pick the right technology, you have to ask yourself the following four questions. First key question is, what's my application? Know where you are today and where you want to go in the next five years, and understand how technology helps solve your business problem over the next five years. Ask yourself, what are my bandwidth data plan, latency, and availability needs? Is my data transferred in burst or steady over time? Do I need to transfer a few bytes or several gigabytes per day? Do I need the date in real time, or are several seconds, minutes, or maybe even hours okay? Do I need connectivity on a campus, in a city, nationwide, or worldwide?
The next question that you want to ask yourself is, what are your environmental, power, and connectivity needs? Is your application in an AC-controlled environment or outside in Alaska or Arizona? Do you have AC power or do you need to rely on solar batteries? Do you need to be always connected? Does downtime mean a loss of revenue or penalties?
The last question you should ask yourself is, should I wait for 5G? Do you want to invest in bleeding-edge technology like 5G or would you rather use proven leading-edge technology like 4G? Does bleeding-edge technology make your product or services any better? And remember that 4G is evolving to 5G, so avoid the technology trap here.
With these questions in mind, let's look at what to transition to. And today, we're really at a fork in the road. If we turn left, we can leverage Gigabit LTE for high-speed applications. If we turn right, we can leverage 4G LTE optimized for IoT applications.
So let's turn right to the next slide. Let's take a deeper look at the 4G LTE technology options for IoT devices. First option is CAT3 and CAT4. With theoretical speeds up to 100 to 150 megabytes per second, CAT3 or CAT4 is great for IoT routers connecting multiple devices. It's overkill for most single device IoT applications.
The next option is Category 1 with speeds up to 10 megabits down, 5 megabits up. It's a good fit for many IoT applications that require mains power. Example, digital signage, retail kiosks, ATMs, industrial controllers, or security cameras. CAT1 was defined in 2008. Like CAT3 and CAT4, it is globally available where LTE is available. So in general, CAT3 and CAT4 and CAT1 are ubiquitous.
The next category, CAT-M, or also known as LTE-M, is a good fit for traditional 2G IoT applications as well as battery-powered IoT sensors. CAT-M was defined in 2016 and is not globally available yet. Predominantly, it's available in markets with early LTE adoption. For example, North America, Latin America, and parts of Asia. NB-IoT, also known as Narrowband-IoT is a good fit for battery-powered devices such as sensors. Like LTE-M, it was defined in 2016, and it's also not globally available yet. It is predominant in markets with late LTE adoption. For example, Europe. One difference of NB-IoT compared to the three other options is that NB-IoT does not support true mobility. So that doesn't mean that your device cannot move, but you will not be able to roam from tower to tower with NB-IoT, but you could disconnect from one tower and connect to the next tower.
So in summary, there are multiple options available for 4G LTE for IoT. Each technology has its pros and cons. To make things more complicated, Carousel also considering to rollout LTE-M or NB-IoT as a secondary network.
Now, let's take a left turn and look at Gigabit LTE and the 4G evolution to 5G. The Third Generation Partnership Project or 3GPP is a collaboration between groups of telecommunications standards associations. 3GPP defines the standards that built the foundation of cellular networks such as LTE. LTE stands for Long-Term Evolution. Since its initial release in 2008, LTE has evolved and is continuing to evolve every year towards 5G. Typically, 3GPP releases a major update of the standards every three years followed by a minor release. To differentiate between major LTE releases, 3GPP has introduced marketing names such as LTE Advanced and LTE Advanced Pro. Release 13 was a key milestone for Gigabit LTE. Release 15, to be released later this year, will be the first standard defining 5G.
Now, let's look at the ingredients for Gigabit LTE in more detail. In order to achieve Gigabit LTE speeds, we need four key ingredients: more RF channels, higher order modulation, more antennas, and more spectrum. Let's look at each ingredient in detail. The first ingredient is more RF channels, also known as carrier aggregation. Think of this as multiple highways to transport more data. Carrier aggregation combines bandwidth from multiple cellular bands of the same mobile network operator. This provides higher peak data rates, better spectral efficiency, and more capacity for the burst usage. This is important to note, as many mobile network operators do not have 20 megahertz of license spectrum per band available.
The next ingredient is higher order modulation. You also sometimes hear this referred to as 256 QAM. QAM stands for Quadrature Amplitude Modulation, and it's widely used in digital data radio communications. Higher order modulation uses more bits per symbol to increase spectral efficiency. This results in shorter transmissions, higher speeds, and higher network capacity. What this means is you can transmit more data at a given point in time.
The third ingredient is MIMO, which stands for Multiple Input, Multiple Output. MIMO uses multiple antennas to transmit and receive data in parallel. Think of this as a multi-lane highway where you can have more cars going in parallel. Most cellular devices today have two antennas per cellular modem. Gigabit LTE devices will require four antennas to achieve higher speeds. For many devices, this means moving from directly attached to cabled antennas.
The fourth and final ingredient is more spectrum. And this is license spectrum which, historically, has been exclusively used for LTE communication as well as shared and unlicensed spectrum. In order to get the additional bandwidth needed for higher speeds, the ecosystem is looking at shared spectrum. For example, in the 2.3 and 3.5 gigahertz bands as well as unlicensed spectrum, which is a spectrum that is shared use. For example, in the 2.4 gigahertz and 5 gigahertz bands, which are, today, used primarily by WiFi and other short-range communication technologies.
So, why unlicensed spectrum? Well, here's a little-known secret. Only very few carriers have enough licensed spectrum available to offer Gigabit LTE with licensed spectrum alone. For example, AT&T has rolled out three carrier aggregation with 4x4 MIMO today offering around 30 megahertz of license bandwidth, which is half of the bandwidth required for a Gigabit LTE. Even with five carrier aggregation, you're only getting 50 megahertz, which is not enough. You can see on this slide, if you use 20 megahertz of licensed bandwidth plus license-assisted access, so some unlicensed spectrum, 64% of the carriers could offer a Gigabit LTE, and then if you only using 10 megahertz of licensed spectrum plus license-assisted access or unlicensed, then over 90% of the carrier's mobile network operators would be able to offer a Gigabit LTE today.
In summary, 4G LTE Advance Pro, also known as Gigabit LTE, is here today and is paving the way to 5G. We need four key ingredients for that to happen. We need more RF channels, or carrier aggregation, we need higher order modulation, we need more antennas, also known as MIMO, and we need more spectrum both in the unlicensed and licensed bands. One thing worth noting is you will not see Gigabit LTE speeds right away. Expect speeds above 100 megabits per second on licensed LTE networks alone under good conditions. And you can expect higher speeds where unlicensed spectrum is available and also where the carrier infrastructure is available.
Now, let's look further into the future and talk more about 5G. When we talk about the 5G evolution, we have to look at two different tracks. 4G LTE, as we know it today, will eventually evolve into 5G LTE. 5G New Radio or 5GNR will define a new radio interface and new spectrum and compliment 4G LTE and will enable new use cases. 5GNR means a massive investment by the mobile network operators. John Legere, T-Mobile CEO, estimated that a nationwide 5GNR network in the U.S. will cost each carrier 1.5 trillion U.S. dollars. So this is a massive investment and a massive deployment that will not happen overnight.
Let's look at the challenges in the context of the spectrum that 5GNR will be using. Below six gigahertz or Sub-6, the wall/building penetration is similar to 4G LTE and WiFi today. The radio and network access is also very similar. However, on 5GNR, or also known as millimeter wave, that is very different. So we're talking about spectrum above 24 gigahertz, and at those frequencies, there is little to no wall or building penetration and we require new infrastructure both for the antennas and for the devices. So, looking at antennas today, infrastructure antennas are typically mounted on towers radiating horizontally. With 5GNR, these infrastructure antennas will be mounted either high or low, radiating vertically. So think of these as shower antennas mounted on tall buildings or lighting polls radiating down or like sprinklers, antennas mounted in manhole covers radiating up. Also, on the device side, antennas today are attached with RF cables going to an integrated radio. With 5GNR, you will have the radio and the antenna integrated together and you will have an electrical wire or fiber between the device and the antenna. One thing to note is that different carriers focus on different applications and different spectrums. So for example, AT&T has publicly announced that they are focusing on mobile applications, whereas Verizon is, at the moment, focusing on fixed 5GNR applications.
When do you need to start thinking about 5GNR? In 2018, 3GPP will finalize Release 15 and network trials will begin. First generation chipsets will become available and the first non-stand-alone trials will begin. Non-stand-alone means that there is a 4G LTE connectivity and then there's also a 5GNR connectivity in parallel. In 2019, 3GPP will finalize Release 16, and network roll-outs will begin. We expect to see devices based on first-generation chipsets such as mobile hotspots or fixed wireless devices at the beginning of the year, and then second generation chipsets towards the end of the year. In 2020, network roll-outs will continue and we expect to see critical mass in urban environments and devices based on second generation chipsets such as handsets. We also expect to see first generation 5GNR modems or modules in the 2020 timeframe. Then beyond 2020, so 2021 and later, the mobile network operators will continue with their build-out, and our prediction right now is that we'll see nationwide 5G by 2025. Also in 2021, we expect to see the first 5GNR commercial routers to become available.
In summary, 5GNR offers many challenges and opportunities. We see a lot of challenges around signal propagation and infrastructure cost, but we also see opportunities in new applications especially around ultra-reliable, low latency, and fixed wireless. Beware of carriers marketing 5G available in 2018, read the fine print. We expect to see first commercial networks in 2019, 2020, and a wider nationwide deployment and build-out by 2025. Remember that 4G evolves to 5G. Over the next few years, expect significant speed and latency improvements along the way. But you will also need to consider low latency on the radio access network versus the end-to-end application latency in this case. So that means that your application needs to move closer to the edge, also known as edge computing.
I want you to remember these key takeaways. First, determine your application needs by asking the key questions. Don't get left behind with 2G and 3G technologies, shutdowns are imminent. Don't be too early with 5GNR, wait until the dust has settled. Leverage 4G LTE technology today from low speed, low power, LTE for IoT, to high-speed Gigabit LTE. Digi has the right products for your mission-critical applications available and we're here to help.
This concludes my presentation on how you can leverage LTE today for your applications and prepare for 5G. I hope you found this webinar useful. If you have any questions, please reach out to me. For additional information, check out our website and connect with us on social media. Thank you very much and have a great day.