Empirical formula of effect on channel estimation error for adaptive MIMO-OFDM based WLAN system

Abstract

The IEEE802.11a standard for indoor wireless LAN systems was released commercially and is wide spread in the marketplace. To achieve higher throughput, IEEE802.11n was proposed and is now undergoing standardization. The MIMO-OFDM technique, which employs multiple antennas for transmission and reception, is the core technology for IEEE802.11n (Foshini and Gans, 1998). This technique requires estimation of channel state information (CSI) on the receiving side. Using the technique called adaptive MIMO-OFDM (Telatar, 1999) in which CSI is applied to both the transmitting and receiving sides, an even higher throughput is achieved. To achieve the maximum performance from adaptive MIMO-OFDM, we must accurately estimate the CSI for all the subcarriers; however, numerous calculations are needed to obtain this estimation. So, to actualize the adaptive MIMO-OFDM system, the calculation load must be decreased. This can be accomplished by applying one CSI to other subcarriers. However, channel error exists between the actual channel and the applied CSI. The channel error also plays a role in decreasing the communications quality (Medard, 2000). The communications quality when using the adaptive MIMO-OFDM technique that has CSI error was evaluated by simulation (Narula et al., 1998). However, there are few measurement evaluation reports on CSI error. This paper focuses on the ratio between the power of the eigenvalue and the interference power that suppresses the communication quality based on the measurement results. We propose an empirical formula for predicting the signal to interference ratio (SIR), which is the ratio between the power of the eigenvalue and the interference power. This SIR empirical formula helps to predict the communication quality and simplify the simulation.