Multiple antenna systems provide high spectral efficiency in rich scattering environments. However when there are a large number of transmit or receive antennas, the complexity also the high cost of mutiple power amplifiers and low noise receivers make them less appealing for practical purposes. One way to deal with this issue is via antenna subset selection. It has been shown before that the use of subset selection at transmit side (which requires a very low rate feedback from transmitter to receiver) can considerably improve the data rate in a  MIMO channel, however, one major problem with this technique is the issue of computational complexity. Optimal subset selection has a complexity which grows exponentially with the number of antennas, hence it is not practical. In this presentation we propose an algorithm which leads to almost the same ergodic capacity as that of an optimal selection and its complexity grows only linearly with the number of transmit antennas. We also provide an analytic upper bound for the capacity of the selected subsystem which is asymptotically tight.