Fifty years from now, when you’re aged and wrinkled (assuming we won’t have discovered the cure for aging) you’ll be sure to tell your grand-kids of the war for the rise of the 5G cellular network from 2016 – 2020. You’ll explain to them how the differences in spectrum efficiency, power efficiency and performance in connectivity were used to decide the ultimate victor. The challengers being MU-MIMO, D2D, NOMA and mmWave technologies. How the battles were fought, not on the ground or sea, but through the electromagnetic spectrum led by top research scientists all over the globe. You’ll remember to tell them that the puppet-masters were the telcos giants while us, the wee subscribers, could only hope for the best outcome. Better yet, direct them here for a simplified description of the technologies currently involved in the 5G revolution.
Silly intros aside. I recently attended a lecture on Key Wireless Access Technologies in 5G and IoT Systems held on 15th January 2018 at Strathmore University’s Transcentury Auditorium. The event was organized by IEEE ComSoc in conjunction with the university. The IEEE Distinguished Prof. Rose Qingyang Hu, Figure 1, delivered the lecture on her ongoing research on the next generation of wireless communication. Her research focuses on the network design and optimization schemes, the Internet of Things, cloud and fog computing, multimedia QoS/QoE, wireless system modelling and performance analysis. In this post I share my insights on what I learned.
Just as in the previous fourth generation, 4G, the search for a stable standard to be used for cellular communication involves a conflict between newly discovered or improved technologies. 4G technologies included LTE (Long Term Evolution), WiMAX (Worldwide Inter-operability for Microwave Access) and UMB (Ultra Mobile Broadband). As we now know, LTE has become the preferred 4G technology and has been widely implemented and accepted. In Nairobi, increased 4G coverage has been going on for about 2 years now. However, as user requirements have increased, a new standard of communication needs to be established. 5G promises to be the answer to the limitations with the current modes of communication. The 5G model of communication especially focuses on handling IoT (Internet of Things) devices. IoT devices broadly refer to the increasingly ubiquitous devices connected to the internet for general (smartphones, tablets) or specific (autonomous cars, wearables) applications. A quick example would be a smart stapler that connects to an app on your phone to tell you how many staples are left, an extremely useful device in my opinion. Let’s explore these competing technologies in 5G.
MU-MIMO
This stands for multi-user – multiple input, multiple output. MIMO systems are communication systems involving signal transmission with multiple antennas at the source and multiple antennas at the destination. MIMO systems have been in use in existing 4G networks, however to handle the requirements of 5G, MU-MIMO has been proposed to bring additional improvements. In MU-MIMO, MIMO is performed simultaneously on n number of user equipment (UEs) as in Figure 2. In MIMO we have a channel matrix H for M transmit antennas and N receivers.
In order for this to be possible, beamforming, which involves space division beam direction, separates the receivers. In 4G this exists but its performance is limited. 5G will be set to push this to its limit, however it is only easy said as some factors make it challenging to even consider. The factors include:
- Number of transmitting and receiving antennas
- Coverage and the number of UEs to be supported per antenna
- Precoding scheme to be used
Precoding schemes involve a matrix ‘code’ that are sent with the message so that the receiver can achieve ‘pre-knowledge’ of the channel. Useful especially in MU-MIMO systems to reduce corruption as well as optimize message reception by the receivers. Such schemes include matched filter precoding, zero-forcing (ZF) precoding, and transmit Wiener precoding.
The Signal to Interference plus Noise ratio for the ith user served by one MU-MIMO base transceiver is:
![\[ SNIR_i = \frac {S} {I+N} = \frac {||w_i h_i||^2} {\sum_{j=1\neq i}^{N} {||w_j h_j||^2 + \sigma_i^2 } } \]](https://i2.wp.com/symonmk.com/wp-content/ql-cache/quicklatex.com-fa9a489fec5038c8c0fa80d5bc3a67e4_l3.png?resize=324%2C50 "Rendered by QuickLaTeX.com")
where:
is the precoding vector for the ith user
is the channel vector for the ith user
is the variance of the complex circular zero mean white Gaussian noise at the nth user
There’s a great YouTube lecture on MIMO and MU-MIMO calculus.
D2D
Device-to-Device communication is another technology associated with the rise of 5G networks. It basically means the direct communication of devices with each other. UEs involved in D2D will have to be at close proximity so as radio communication can be sustained. This will greatly reduce the load on the base station transceivers or access points. Existing technologies already enhance this form of communication such as Bluetooth and WiFi-Direct. It is up to cellular networks to leverage such concepts for the increasing number of communicating devices. Devices use existing cellular infrastructure while switching to D2D when location proximity is sufficient as in Figure 3.
Researchers have deduced that there are several advantages to this scheme including:
- Ultra-low latency in communication as the signal path is greatly minimized
- Load on core network is reduced, increasing spectral efficiency
- Supports a variety of emerging applications such as Machine-to-Machine (M2M) and context aware applications
D2D communication has already been established as part of the 4G-LTE standard defined by the Third Generation Partnership Project (3GPP) Release 12.
D2D looks very promising as some of its features have been in use for quite a while now. Some challenges, especially in security and pricing for service providers are yet to be fully solved.
NOMA
Non-orthogonal multiple access (NOMA) schemes have become recently popular in achieving spectral efficiency, becoming attractive for the 5G networks. Currently up to 4G, cellular networks have been using orthogonal multiple access (OMA) schemes. They include the frequency division multiple access (FDMA), time division multiple access (TDMA) and code division multiple access (CDMA). NOMA has been recently proposed as it meets the requirements of 5G networks unattainable using OMA.
NOMA can be achieved by two techniques of power-domain and code-domain. In the power-domain NOMA scheme, superposition coding at the transmitter and successive interference cancellation (SIC) at the receiver allows for multiplexing of UEs in the power domain utilizing the same spectrum. Superposition coding is performed on the signals where they are superimposed into a single waveform and passed through the transmitter. At the receiver the SIC decodes the incoming signals through subtraction based on their power levels until all the superimposed signals have been separated. In Figure 4, the strong UE’s (User 2) signal is first decoded, while the weak UE (User 1) may be assigned more power first to guarantee fairness. The system throughput of this technique is significantly better as several UE’s can efficiently utilize the same bandwidth resources. This is especially useful for IoT devices as the improved spectral efficiency will enable handling of more simultaneous connected devices.
The limitations with this scheme are:
- Increased receiver complexity in order to carry out the SIC operation
- Higher energy consumption compared to OMA, since each UE has to decode signals of all other weak UEs in the same cluster
- A considerable difference in the signal powers between two UEs should be there to take advantage of the power domain scheme
NOMA’s unique advantage has made it to seriously be considered for the 5G cellular network standards. NOMA can also be used alongside other technologies such as MIMO-NOMA and mmWave-NOMA. Beyond 5G, NOMA is already been used for TV broadcasting.
mmWAVE
The millimeter wave mobile communication has gained interest over the last few years for use in 5G. Upcoming technologies such as VR, AR, 4K video and increased connected devices may benefit from higher bandwidth networks and here is where mmWave communication comes in. The bandwidth in the mmWave is from 30GHz to 300GHz. In comparison, current mobile communication bands are on 850/900/1800/1900MHz bands with Bluetooth and Wi-Fi mostly on the 2.4GHz and 5GHz bands. However, the mmWave method immediately faces many challenges:
- Changes in the physical medium, the media access control (MAC) and routing layers to enable mmWave communication
- Propagation can be achieved over short distances (few kilometers) as it experiences high propagation losses through the atmosphere, rain and obstacles like buildings, humans, furniture.
Several techniques have been suggested to work alongside mmWave bands to counter its existing limitations. Beamforming as discussed earlier could be used to provide directional propagation of signals. Directivity allows for further exploitation of the transmission using spatial reuse.
mmWave communication can be collaboratively used alongside MIMO and NOMA schemes discussed earlier.
These are some of the key technologies Prof. Rose mentioned in her lecture alongside the detailed mathematical expressions and concepts supporting each technology. It was very interesting to become aware of the amount of work researchers are doing to take us to the next level of communication. Come 2020, hopefully we’ll be on 5G.