New demands in wireless communications imply an important increase in channel capacity and thoughput. In recent years, research in wireless communications has focused mainly in the utilization of spatial and temporal diversity in both transmission and reception by means of antenna arrays at both ends, in order to achieve this increase in channel capacity. These MIMO (Multiple-Input Multiple-Output) systems, to work properly and guarantee its potential advantages, demand powerful signal processing procedures so as to recover the signal transmitted by the antenna arrays in the best possible way. Among the different signal processing algorithms that exist for these MIMO systems, the BLAST algorithm has come to be a potential alternative for its excellent complexity-performance tradeoff. This system has different realizations. In this paper, we concentrate on a Least-Squares based VBLAST architecture and propose a simple means for improving further its performance. Indeed, with the concatenation of a conventional convolutional coder, the overall performance can be increased in 3 dB while increasing the complexity in an acceptable way. Results are also compared to the quasioptimal Closest Point (CP) decoder. It is shown that our proposed system outperforms the CP algorithm by 3 dB for the same antenna arrays, and for the same computational complexity, our proposed system is better by 4 dB.