We consider a MU-MIMO downlink channel which comprises a base station (BS) and $ K $ mobile stations (MSs). The BS has multiple transmit antennas of $ M $ elements while each MS has a single antenna. By the exploitation of multiple transmit antennas in a multiple user environment, the network is able to provide high through performance. Moreover, the required complexity for each mobile is much lower than the single user MIMO scheme, which mandates each mobile to have more than one antenna.

## MU-MIMO with feedback signalingEdit

Let us discuss the system model to evaluate the performance of MU-MIMO especially feedback signaling is used for the operation of a MU-MIMO system. We model downlink data channels from the BS to MS and uplink feedback channels from each MS to the BS. Throughout this paper, we consider MS 1 for the performance evaulation without loss of generality. We denote the distance between the BS and MS 1 as $ D_{b,m} $ and denote the received powers of MS 1's downlink data channel and MS 1's uplink feedback channel as $ P_{b,m} = p_b D^{-2 \gamma}_{b,m} $ and $ P_{m,b} = p_m D^{-2 \gamma}_{b,m} $ respectively where $ p_b $ and $ p_m $ are the transmit power of the BS and MS 1 respectively.

We assume that $ \mathbf{x}_{b} $ and $ x_{m} $ are transmision signals from BS and MS, respectively. MS 1's downlink data channel is modeled as:

- $ y_{b,m} = \sqrt{P_{b,m}} \mathbf{h}_{b,m}^H \mathbf{x}_{b} + n_{b,m}, $

and MS 1's uplink feedback channel is modeled as:

- $ y_{m,b} = \sqrt{P_{m,b}} h_{m,b} x_{m} + n_{m,b}. $

where $ n_{b,m} $ and $ n_{m,b} $ are noise signals which are assumed to be a complex Gaussian iid random variable with zero mean and unit variance at both the BS and MS 1, respectively.