MIMO (Multiple Input Multiple Output)

The dream of every communication designer is to provide good quality service to each of his customer within the limited available bandwidth. But in the recent years, the amount of data flowing through the channel has increased. But there has been no such high increase in the available bandwidth. So it is a challenge to effectively utilize the channel spectrum along with providing good quality of service across wireless links. One of the possible ways is to use multiple antennas at both the ends of the transmission link. MIMP exploits natural phenomena like multi-path propagation to increase throughput, reduce error rates etc. rather than trying to eliminate them.

The main motive behind the MIMO was to increase the user data rates within the constrained spectrum. The initial application of MIMO was proposed for indoor wireless LAN, fixed wireless access networks etc. But now the aim to widen its applications to cover other areas also. The core idea behind the system is to exploit the de-correlation of multiple received signals in the presence of multi-path propagation, there by separating the data streams occupying same bandwidth. Hence Rayleigh fading and constrained total power comes into the figure.

Data is transmitted in busts and the receiver knows the channel pattern through the usage of training algorithms; but not necessarily the transmitter. The training sequence enables the receiver to predict the channel coefficients and extract the data coming in multiple streams. To maximize the effect of transmitting rate half of the time interval is used for data transmission and half for training. The adaptive transmission is possible only if the transmitter knows the channel coefficients in advance, so that the more data can be sent through good channels. In case of time division duplex channels, this requires the channel to be stationary and hence channel details need to be fed back at the same rate as the channel characteristics are changing. In case of the frequency division duplex channels, the coefficients should be transmitted at a different frequency. To overcome the fast feedback requirement, the spatial mean of channel coefficients has been proposed; instead of the instantaneous values. This greatly enhances the channel capacity, in case of correlated channels.

There is an ongoing effort to standardize the MIMO standard under the name IEEE 802.11n. It will offer up to eight times coverage and about six times data rates, of current 802.11g networks.