Parent Category: 2021 HFE
By Brendon McHugh
Mobile communications companies worldwide are just in the beginning of developing mmWave based telecommunication networks. A lack of spectrum in the legacy (sub-6 GHz) radio bands has demanded the use of the mmWave and sub-THz frequencies to fulfill the growing wireless data demands from an increasing number of Internet connected mobile devices.
Next-generation networks must address not only the growing user and data bandwidth needs but also provide service for new applications including a growing list of critical applications; from autonomous driving to remote surgery, to infrastructure sensor and control systems. Moreover, these networks (both licensed and unlicensed) must provide reliable coverage for users in dense urban, and sparse rural, areas using a variety of interconnected nodes making use of deployment models such as mesh networking, small-cell/macro-cells and device-to-device (D2D).
This article discusses the use of multi-user multiple-input multiple-output (MIMO) and software defined radio (SDR) technology allowing for 5G networks to provide a high level of service efficiency across different environments.
To see how 5G networks will enable the massive amount of bandwidth, we must first look at the fundamentals of digital communications over one channel; the Shannon-Hartley theorem:
The maximum achievable bit-rate (with arbitrarily low bit-error rate (BER) and without error correction) is referred to as the channel capacity, C:
Equation 1: the Shannon-Hartley Theorem.
What Equation 1 tells us is that, to increase the capacity, C, we can increase the bandwidth, increase received power, or decrease the noise on the channel. That’s why with 5G, a lot of focus is on increasing the density and diversity of the network by using small cells or alternative infrastructure. This leads to a closer signal source, and thus a higher signal strength, so therefore a greater the signal-to-noise ratio.
As well, 5G increase bandwidth, B, by using the legacy sub-6 – 7 GHz bands as well as bands from ~24 – 52 GHz, in which spectrum is often auctioned off and regulated by governments. Moreover, by allowing for heterogeneous access to the 5G core network, it allows for a greater diversity of devices to act as base stations and use frequencies beyond 6GHz. This, as well as alternative modulation schemes, such as the New Radio (NR) 5G standard is based on OFDM (Orthogonal frequency-division multiplexing), a method of modulating a digital signal across several different channels; reducing the overall interference, and thus noise, while also increasing bandwidth. Overall, then, channel capacity is increased.
Challenges for 5G Networks
In free space telecommunications, a simplified model, called the Friis equation (Equation 2), describes the free-space path loss (FSPL) the available power at the receiver, while accounting for the system loss factors:
Equation 2 • the Friis equation.
As we will see, much of 5G is resulting from economic and practical matters. Increasing bandwidth often requires a lot of regulatory red-tape and noise is fundamentally beyond our control. Moreover, Equation 1 contains an important factor, the available power at the receiver, which increases capacity, as shown in Equation 2. We see that power at the receiver is increased/decreased through various factors, often which are beyond our control (loss factor, frequency) and also highly regulated (transmit power, distance between transmitter and receiver). Thus, well will see that this is why a lot of focus has been on massive MIMO antenna design.
The mmWave frequencies (between 30 and 300 GHz) enabling these technologies possess unique physical limitations and challenges for the telecommunications industry. Microwaves suffer from degraded FSPL communication, because mmWave, unlike lower frequencies, are not bent or refracted by the ionosphere and are strongly attenuated by certain dense materials such as concrete walls in buildings. Much of this stems from mmWave propagation characteristics contributing to path loss, including diffraction and blockage, rain attenuation, atmospheric absorption, and foliage loss behaviors, which we will call system loss factors. A number of challenges come from these.
From Equation 2, we can see that overall power at the receiver decreases as the frequency is increased (since c=fλ), decreases with an increase in distance to transmitter, as well as with system loss factors. The loss factor (fudge factor), L, includes various system loss factors that come into play when considering real-world conditions of path loss and considering link budget/margin requirements. This includes the following broad categories of which numerous mathematical models have been developed (of course, many other models exist for path loss in microwave cellular networks including Okumara-Hata model and the alpha-beta-gamma (ABG) model):
1. Diffraction: the electromagnetic (EM) beam wavelength is much smaller; mmWave communication signals do not diffract as strongly and thus are more susceptible to blocking by physical objects (such as those found in urban areas) than sub-6 GHz signals used in older generation communications networks
2. Rain/shadowing: rainfall/precipitation attenuation is a phenomenon noticed relative to the rainfall/precipitation rate and frequency. This results in increased path loss, limited coverage area, and consequently degrading the system performance -- all in all, reducing quality of service (QoS)
3. Foliage and ground reflection loss: signal attenuation occurs through foliage or reflections off of ground substances
4. Atmospheric absorption: damping/attenuation occurs due to atmospheric chemistry and because at mmWave frequencies, energy is absorbed by oxygen and water vapour (and other gases). The effect of water and oxygen on signal attenuation is shown in Figure 1 below.
Figure 1 • atmospheric losses from absorption of mmWave frequencies.
All of these must be factored into the link budget/margin, and can also change depending on season, weather, or other natural and man-made environmental condition (after all, trees don’t have leaves all year round and it only rains sometimes).
Not only do the physics of EM waves prove to be challenging, but also the networking infrastructure at all levels; from the user device to small base stations to security to actual large scale base station antenna/tower design. A lot difficultly lies in the amount of base stations that must seamlessly work together; imagine making a phone call in your house over a WIFI router, getting into your car and leaving the range of the router, going into a building with it’s own D2D communication, and still remaining on the same call without and service disruptions: the problem is quite challenging.
Since 5G relies on point-to-point communication, the path loss is significant for mmWave, networking between various nodes is quite challenging. 5G provides for the use of micro- and/or small- cells to provide diversity and ensure users have coverage, as opposed to macrocells (cell towers) often used today. As discussed above, almost any object in the line-of-sight (LOS) between mmWave small cell transceivers, will attenuate the link/signal strength between those two cells.
New and existing base stations avoid interference and signal interruptions by continuously tracking receiving and/or user equipment (UE) and then recalculating optimum data paths between cells. This enables transmission and receiver adjustments in real time to ensure consistent, uninterrupted data flowing to and from 5G devices (for example, vehicles or cell phones) that are on the move or when objects block a signal signal path while also accommodating for UE power limitations (battery life vs. transmit power). To accomplish much of this, often beamforming and beam steering is used.
A phased-array antenna allows for beam steering and optimization of the radiation beam to maximize the peak effective radiated power (ERP) towards a mobile device receiver within the antenna’s cell sector. This is important as ensuring sufficient gain is problematic for mmWave radio communication, since there is much lower power amplifier efficiency in the 24 to 40 GHz bands. To combat this, engineers have turned to massive MIMO architectures with at least 64 antenna radiating elements. Utilizing antenna arrays for multi-beam links allows for the low transmitter power to achieve linear ERP but high gain in the range of 60 dBm. Again, this will contribute to an increase in channel capacity, C.
The use of multiple spatial data streams (i.e. multi-user access), shown in Figure 2, using beamsteering and beam forming while supporting dual polarization at the transmitter and receiver, will also contribute to channel capacity due to the MIMO nature of phased-arrays.
Figure 2 • Small and Macro Cell Base Stations with MIMO beamforming capabilities within a 5G network.
To see how capacity of such systems are formed, we will take as an example a multi-user MIMO base station receiver, which we can theoretically model in Equation 3.
Equation 3 • Multi-user MIMO uplink receiver system.
From Equation 3, we see that there are notable differences compared to the one channel Shannon-Hartley theorem or even single-input multiple-output (SIMO) systems. These are:
- Users of the K antennas do not cooperate, i.e. they are independent signals.
- There are now K capacities that form up a the complete “sum capacity” of the system
- Each user has it’s own power budget subject to the power constraint outlined in the power limitation expectation value above.
From this, and without going into too much detail, we are able to model our theoretical MIMO multi-user uplink receiver capacity as a “sum capacity”, as shown here:
Equation 4 • Sum capacity of a multi-user MIMO uplink receiver system.
On the other side of the coin, we can look at base station transmit capabilities. To accomplish spatial multiplexing from a base station, the use of multi-user/multi access channel MIMO linear antenna array structures within network cell sectors is paramount. Beamforming is a calculation-intensive process that requires active MIMO phased-array antennas within network cells with robust signal processing capabilities. The small wavelengths of mmWave frequencies enable large numbers of antenna radiating elements to be deployed in the same form factor thereby providing high spatial processing/multiplexing gains that can theoretically compensate for at least the some path loss. Path loss can be mitigated through beamforming gain which results in ray-like beams for either point-to-point or point-to-multi-point capabilities. The ability to rapidly steer multiple independent and sharp beams is critical for the robustness and channel capacity in mmWave network systems.
A well-designed phased-array antenna for 5G should factor in dual polarization, minimal array size, mitigating side lobe level (SLL), optimal beam steering angle range/resolution, minimal system noise, power efficiency, and high gain to compensate for propagation path losses, to name a few. Also of importance is the ability of the antenna array to steer beams from point-to-point (high-gain) and point-to-multi-point (wide-angle/dimensional) to/from user devices and other network nodes (so called “spatial multiplexing”).
This increase in 5G small cell sector networking also comes with increased security issues. Since 5G wireless networks works well in the presence of smaller antennas and base stations located indoors and outdoors, devices and the network are particularly vulnerable to security hacks. Information on a 5G cell tower / antenna can reveal a mobile user’s location and even what building a user is present in, as the user communicates with the antenna each time. This data can lead to threats like semantic information attacks to cause harm. Access point algorithms in 5G mobile networks can also leak location data. Thus, more 5G antennas permit for precise location tracking of users inside and outside. For example, security researchers in 2018 found faults in the 5G Authentication and Key Agreement (AKA) security protocol that could potentially be used to steal sensitive information; including personal banking and credit card information.
To conclude, it is evident that continuing experimental research on redesigning all layers of the 5G wireless protocol stack and network infrastructure will need to take place in order for the full realization of mmWave technology; this includes the physical layer, MAC layer, network and transport layer, applications layer, security, devices/antennas and materials.
Applications Requiring 5G
1. Vehicular communications: 5G networks will make autonomous vehicles much more popular. The ability of 5G to support many simultaneous connections per unit area compared to older networks and at very low latency and D2D communications is particularly important for this application. The following capabilities are expected from this technology:
- Centrally collecting data from moving cars and providing aggregated information, including real-time maps, traffic data, road hazard warnings, and driving recommendations.
- Data collected will enhance car and driver safety and provide planning capabilities and improve road usage efficiency, all through coordinated driving.
- Vehicle-to-network (V2N) communication will enable other applications not directly related to autonomous driving, such as self driving/collision avoidance, preventive maintenance, and access to multimedia entertainment—all of which demand substantial bandwidth.
2. IoT devices: the promise of this technology, for many industries, arises from 5G networking, which will again provide the low latency, high bandwidth, D2D and algorithms for various signal processing, such as data compression schemes. A few interesting applications worth noting are:
- From industrial automation/factory monitoring to robot and drone delivery systems to remote surgery and diagnosis; a wide variety of connected devices comprising the internet of things (IoT) revolution is underway.
- IoT revenues will start to be generated from platforms, apps, and services. Operators of IoT will also be able to manage various service partners, refine data from their platforms and turn big data into smart data using AI and machine learning.
3.Mobile phones: unprecedented data speeds and low latency links will provide a number of capabilities to mobile device segments, including:
- Virtual reality (VR) and augmented reality (AR) on mobile phones that can simulate anything and provide an immersive experience for users.
- Holographic/Mixed reality (MR); from holographic conferences to attending work meetings.
- Connections to cloud computing services which are expected to increase alongside data upload/download.
- Increased use of mobile phones for internet connectivity instead of fibre optic cables, etc.
- How software defined radio (SDR) can help the 5G industry resolve challenges
Software defined radio provides a number of advantages to aiding in the design of 5G networks and resolving problems related to networks. High performance SDRs are very modular and consist of commercial off the shelf (COTS) products, which are capable of standing the test of time through the ease of swapping out parts. Moreover, an SDR is very flexible in it’s ability to be upgraded to new firmware and software for the latest communication protocols and upgrades in capabilities. This is particularly important as new 5G frequency bands come online through 3GPP. SDR also benefits from having easy to use software development tools to experiment with real world scenarios, such as mmWave signal path losses, using programs including GNU Radio, a popular open source platform.
Functions related to antenna control (Figure 3) such as beam reconfiguration functions, can be accessed directly and efficiently from the same software interface; including orthogonal phase and gain control during beamforming, changing waveforms and frequencies, modifying the beam shape while maintaining a beam direction, or steering the beam while maintaining a beam shape, as well as polarization of these beams. Also possible are data link quality controls using SDR processing, including error vector magnitude (EVM) and bit error rate (BER). Finally, SDRs can be simultaneously used for monitoring the spectrum as well as base station signals, call/data access rates, speed tests, mobile network latency, and more.
Figure 3: Capabilities of SDRs for antenna control and link metrics.
SDRs are also able to be deployed in high volumes, which are essential to controlling the many antennas and base stations that have to be installed on homes or buildings in order to maintain coverage and channel capacity. Extra repeaters will be required to install in urban areas for proliferation of signals at a wider distance while also maintaining uniform speeds in densely populated areas.
Previous generations of mobile networks from GSM to 3G to LTE have required significant testing in small geographic locations before being deployed globally by 3GPP. The story of 5G technology will be similar; however more technological advancements and testing will be needed to reach the massive throughput, bandwidth and low latency expected from 5G. Test beds are activity being deployed in cities across the world to make the promise of 5G a reality. Examples of test beds underway include 5GUK in the UK and Rutgers University in New Jersey and COSMOS in New York City. All these test beds have one thing in common: the use of SDRs with small cells which make up an overall 5G network.
SDR based antenna nodes in network cell sectors are publicly accessible and enable researchers to log in remotely and run experiments on firmware installed on the devices. This allows for researchers to run experiments in their own labs, with custom topologies and dynamic blockage and mobility scenarios. Researchers can collect data and stream it for such applications as RF propagation modeling. As discussed above, the propagation loss is highly dependent on the physical environment and the sorts of scattering, diffraction and reflections that occur in everything from urbanized parking lots and buildings to fog and other precipitation.
SDRs also provide numerous capabilities to the cell network as well. One promising capability is mesh networks. Due to the many small cells making up 5G networks, there is a great deal of bandwidth capacity such that network cells can be deployed in meshes with self-organizing technology based on AI and machine learning. These simply re-route traffic around a blockage via neighboring cell links. Much of the computation can be carried out in SDRs with system on a chip technology. As well, optimizing for path loss reduction using beam steering algorithms can be accomplished by exploiting electronically steered phased array antennas with dynamically controlled beamforming techniques within an FPGA. From this technology, SDRs controlling mmWave nodes can establish and maintain multiple wireless connections, either point-to-point or point-to-multi-point, or a combination of both, using spatial multiplexing techniques. SDRs also allow for deploying standards as they are developed, such as those developed by 5G NR access technology. This is the global standard for air interface of 5G networks which details the frequency range of the signals, the spacing between the carrier signals, the slot duration, and other parameters. This and much more is possible with SDRs embedded in 5G network infrastructure.
Conclusion
5G technology is promising for a number of applications. With this promise comes a number of challenges needing to be addressed. This includes the actual physics of mmWave frequencies and the network infrastructure enabling low latency, huge bandwidth and massive MIMO device connectivity. SDRs are proving to be a valuable tool for testing, developing and deploying networks throughout the world, and will continue to play a large part due to the re-configurable and flexible nature of the device.
About the Author
Brendon McHugh is the Field Application Engineer and Technical Writer at Per Vices. He holds a degree in theoretical and mathematical physics from the University of Toronto.
http://www.cs.man.ac.uk/~barry/mydocs/CS3282/Notes/DC06_7.pdf
https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-com.2018.5044
https://en.wikipedia.org/wiki/Free-space_path_loss
https://5g-tools.com/qos-for-4g-lte/
https://www.cosmos-lab.org/
https://www.hindawi.com/journals/ijap/2021/5559765/
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwix7__WwtvyAhUCTt8KHSc2AXkQFnoECAYQAQ&url=https%3A%2F%2Fwww.mdpi.com%2F2073-8994%2F10%2F12%2F672%2Fpdf&usg=AOvVaw0AcavFHwGFxmXGOAHD1M5P
https://arxiv.org/pdf/1509.00436.pdf
https://www.edn.com/tiny-waves-big-challenges-getting-5g-mmwave-mobility-right/
https://www.mathworks.com/content/dam/mathworks/ebook/microwave-journal-5g-phased-array-technologies-ebook.pdf