A study on transmit power allocation approach for closed-loop MIMO system

Abstract

Recently, together with a rapid growth of wireless communication service market, users`` demands are extended to wideband and high datarate communication such as high quality multimedia services. To satisfy such demands with limited communication resources like frequency bandwidth, employing of multiple input multiple output (MIMO) sys-tem which utilizes multiple antennas at the transmitter and the receiver is essential. The MIMO systems can be classified into two types: one is space time coding system which improves system error performance by exploiting the spatial diversity, namely, using multiple antennas, and the other is layered space time (LST) system that increases transmission data rate by transmitting multiple data streams over multiple antennas. Among them, vertical Bell labs layered space time (V-BLAST) system is getting momentum due to its capability to support high data rate. In such MIMO systems, increased radio frequency (RF) chains and corresponding complex signal processing are required by employing multiple antennas, and thus these may be important issues from the system implementation point of view. To decode signal from interferences caused by other signals, an accurate signal detection algorithm with high complexity is required at the receiver, especially in V-BLAST systems. Such increased receiver complexity leads to enlargement of mobile terminal size as well as large power consumption. For this reason, huge complexity must be avoided to provide a realistic feasibility of the system. Nevertheless, an increase in computational complexity is inevitable to improve system performance. In conventional open-loop systems, it is well known that maximum likelihood (ML) detection is optimal but has tremendous complexity. For instance, the receiver complexity increases exponentially with the number of transmit antennas, thus, it is hard to practically implement. On the other hand, in closed-loop systems, the system performance can be improved w...