If capacity varies with min of Mt and Mr where those parameters are the transmitting and receiving antennas respectively, then what is the capacity gain from gigantic number of antennas @ base stations, although user equipment at max is equipped with 2 antennas?
There are generally three things that limit the capacity of single-user MIMO transmissions: number of transmit antennas Mt, number of receive antennas Mr, and half the coherence interval. The latter represents the cost of estimating the channels. The following is an early paper on that topic:
L. Zheng and D. Tse, “Communication on the Grassmann manifold: A geometric approach to the noncoherent multipleantenna channel,” IEEE Trans. Inf. Theory, vol. 48, no. 2, pp. 359–383, Feb. 2002.
Massive MIMO in a single isolated cell can be seen as a single-user MIMO system where the antennas at one side are not allowed to cooperate. Hence, it's capacity is upper bounded by the single-cell results by Zheng and Tse.
If you consider a large multi-cell environment, things are different because there is also inter-cell interference. Nevertheless, we observed similar behaviors in our recent paper:
Massive MIMO for Maximal Spectral Efficiency: How Many Users and Pilots Should Be Allocated? (http://arxiv.org/abs/1412.7102)
We show that when the number of BS antennas is large and we can schedule any number of users, the capacity is still limited by half the coherence interval.
The achievable capacity of Massive antenna can be done by using the Massive MIMO as MultiUser MIMO. if we have N users with two antennas, and the base station has Nt antennas, then the capacity is min{Mt, N*Nr}.
There are generally three things that limit the capacity of single-user MIMO transmissions: number of transmit antennas Mt, number of receive antennas Mr, and half the coherence interval. The latter represents the cost of estimating the channels. The following is an early paper on that topic:
L. Zheng and D. Tse, “Communication on the Grassmann manifold: A geometric approach to the noncoherent multipleantenna channel,” IEEE Trans. Inf. Theory, vol. 48, no. 2, pp. 359–383, Feb. 2002.
Massive MIMO in a single isolated cell can be seen as a single-user MIMO system where the antennas at one side are not allowed to cooperate. Hence, it's capacity is upper bounded by the single-cell results by Zheng and Tse.
If you consider a large multi-cell environment, things are different because there is also inter-cell interference. Nevertheless, we observed similar behaviors in our recent paper:
Massive MIMO for Maximal Spectral Efficiency: How Many Users and Pilots Should Be Allocated? (http://arxiv.org/abs/1412.7102)
We show that when the number of BS antennas is large and we can schedule any number of users, the capacity is still limited by half the coherence interval.
Massive MIMO is essentially a 5G cellular technique which shall be realised symbiotically with Millimeter Wave (mmWave) and Small Cell Networks (SCN). Therefore, the capacity shall not be only scaled with number of antennas, but with the increased transmission bandwidth at mmWave, around 500MHz unlike 10-20MHz for current broadband networks, and increased frequency reuse which is natural because of increased path loss at mmWaves. Further the path loss can be compensated by advancements in beamforming algorithms to create steerable sharp beams with certain risk of interference---interference control and mitigation techniques should come into handy here.
Recently, it has been implied to use around 32 antennas on the bottom and top of the user equipment with alternative polarisations, whereas the number of antennas at the base station side can be of the order of 100s. This is spatially permissible at higher mmWave frequencies of 28GHz and 38GHz where the signal dimension becomes much smaller than the electronics in UEs.
Please note that massive MIMO, as it has been proposed and analyzed, is a technology for conventional cellular frequencies (below 10 GHz). These frequencies will be very important also in the future, since they can provide wide-area coverage and support high mobility. Massive MIMO can provide 10x to 50x gains in throughput and uniform user performance in the conventional frequency bands.
In addition, multi-antenna techniques will also be highly useful at mmWave frequencies, but there has not been so much research on this topic. We don't know how the coherence bandwidth/time behave (which fundamentally limits the channel acquisition and multiplexing gains), we don't know how to implement beamforming and precoding in hardware, etc. In other words, we don't know what the achievable gains of massive MIMO are in mmWave. But I encourage people to perform research in this field!
- Badges
- Science topic
- Similar topics
- Engineering
- Telecommunications Engineering
- Antennas and Propagation