A key to the advanced networking future that 5G promises is Massive MIMO. It’s a term you’re likely to hear increasingly about as the industry moves forward with 5G implementation, and it’s one that is worth taking a closer look at.
Massive MIMO (multiple input/multiple output) is all about antennas, radios, and spectrum, all working together to give the kind of speed, capacity, and other benefits businesses are looking forward to in our coming 5G world.
The technologies essential to 5G – Massive MIMO among them – are challenging to support with spectrum below 1 GHz because of the size of the antennas. But when using mid-band and high-band frequencies, the antenna elements become much smaller, enabling more such elements to fit into a given space.
By using a broader range of frequencies, capacity can be increased by using more radios at higher frequencies with smaller antennas. If spectrum is a highway, then MIMO doesn’t just add more lanes; it adds levels, creating multiple-decker highways that vastly increase the capacity of the network.
Traditional MIMO systems have multiple transmit and receive branches (typically up to eight in current systems). However, each branch consists of tens of antenna elements to create high antenna gain needed to boost coverage. Massive MIMO systems, on the other hand, have a much higher number of transmit and receive branches (32, 64, or even higher), with each branch having two to three antenna elements. The high number of antenna branches deliver a huge boost in capacity via the ability to aim the signal to a user both in horizontal and vertical domains.
The capabilities of a Massive MIMO system are determined by the frequencies in use, and the element size of an antenna is dictated by the wavelength of the signal that antenna is built to transmit or receive. Consequently, higher frequency bands with their shorter wavelengths mesh perfectly with the large numbers of antenna elements required for building high‐performance Massive MIMO systems. However, with high frequency bands it becomes difficult to support a large number of Multi-user-MIMO layers.
Frequency bands at 2 GHz or above are the ones best suited for the task, which is why Sprint is in such a strong position when it comes to 5G and the underlying Massive MIMO deployments. We are licensed for much of the 2.5 GHz spectrum, and the fact that we have it in a contiguous block maximizes channel flexibility and enables greater data capacity. The 2.5 GHz spectrum offers the sweet spot for massive MIMO in terms of achievable form factor, capacity and coverage.
Picture a MIMO layer as a dedicated data stream. More MIMO layers result in more throughput. Most recently upgraded LTE systems can support four‐layer MIMO. Massive MIMO systems can support eight or 16 layers currently and will support a higher number of layers – 24 or more – in the near future.
MIMO comes in a couple of different flavors. There is Single-user MIMO (SU-MIMO) in which multiple streams, using multiple antennas and the same spectrum and time resources, are directed to single devices, allowing them to send and receive data simultaneously.
Then there is Multi-user MIMO (MU-MIMO), where multiple streams using multiple antennas and same spectrum and time resources are directed to many devices. Both of these techniques reduce network transmission lag time and make for a more efficient wireless network.
With their huge number of antenna branches, advanced beamforming capabilities, support for higher MIMO layers, and multi‐user MIMO, Massive MIMO systems improve both coverage and capacity in ways that no prior 4G LTE feature has.
We mentioned beamforming. That is the technique used to focus a signal to an intended recipient while minimizing the noise generated by signals meant for other users. Beamforming techniques vary from simple analog beamforming (orienting the entire sector coverage area to where it’s needed) to sophisticated digital beamforming with device‐specific beams.
Newer, advanced implementations use three-dimensional beamforming to steer user‐specific beams precisely. This is particularly useful when there are users both on the ground and inside buildings on different levels.
The results are improved coverage for users at the edge and reduced overall interference.
Time or frequency?
One consideration in 5G networks involves the use of Time Division Duplexing (TDD) or Frequency Division Duplexing (FDD). Both of these approaches provide paths for uplink and downlink traffic.
FDD networks use dedicated spectrum for uplinks and downlinks, and the amount of spectrum for each is generally equal. That doesn’t always fit well with usage patterns; typically there is more data heading downstream than upstream. When equal amounts of spectrum are dedicated to each direction, it can result in inefficient use of that spectrum.
TDD networks, on the other hand, can use the entirety of the spectrum for uplink or downlink. However, they can only transmit or receive at a point in time – basically, the uplink and the downlink take turns – whereas FDD networks can perform simultaneous transmission and reception. But we’re talking milliseconds here, so any difference is something that is transparent to the user.
The use of the same channel for uplink and downlink with TDD allows better channel estimation for beamforming, making it better suited for Massive MIMO systems, in addition to the flexibility it can offer in terms of deployments and efficient spectrum utilization.
Massive MIMO is one element of what is being called 5G New Radio, which describes the radio technology behind all those appealing new features of 5G. You can find out more about MIMO and Massive MIMO, as well as almost anything else you want to know about 5G, in the Ultimate 5G Explainer.