In this paper, the effect of channel estimation errors upon the Zero Forcing (ZF) precoding Multiple Input Multiple Output Broadcast (MIMO BC) systems was studied. Based on the two kinds of Gaussian estimation error m...In this paper, the effect of channel estimation errors upon the Zero Forcing (ZF) precoding Multiple Input Multiple Output Broadcast (MIMO BC) systems was studied. Based on the two kinds of Gaussian estimation error models, the performance analysis is conducted under different power allocation strategies. Analysis and simulation show that if the covariance of channel estimation errors is independent of the received Signal to Noise Ratio (SNR), imperfect channel knowledge deteriorates the sum capacity and the Bit Error Rate (BER) performance severely. However, under the situation of orthogonal training and the Minimum Mean Square Error (MMSE) channel estimation, the sum ca- pacity and BER performance are consistent with those of the perfect Channel State Information (CSI) with only a performance degradation.展开更多
The conventional zero-forcing vector precoding can provide the same SNR for decentralized receiving users.In order to meet different Qos requirement of multimedia applications,we modified the conventional precoding ma...The conventional zero-forcing vector precoding can provide the same SNR for decentralized receiving users.In order to meet different Qos requirement of multimedia applications,we modified the conventional precoding matrix by using a diagonal matrix named relative SNR control matrix so that different SNR can be got for different users＇ requirements.Then,with the new precoding matrix used,we analyze the average SNR of all MTs.We find that if a proper perturbation vector is chosen under the power constraint,the average SNR can be maximized.Finally,we indicate that the choice of perturbation vector is a K-dimensional problem which has low complexity algorithm to solve.Simulation shows that our scheme is very effective for multimedia applications.展开更多
基金by the National Natural Science Foundation of China (No.60496311).
文摘In this paper, the effect of channel estimation errors upon the Zero Forcing (ZF) precoding Multiple Input Multiple Output Broadcast (MIMO BC) systems was studied. Based on the two kinds of Gaussian estimation error models, the performance analysis is conducted under different power allocation strategies. Analysis and simulation show that if the covariance of channel estimation errors is independent of the received Signal to Noise Ratio (SNR), imperfect channel knowledge deteriorates the sum capacity and the Bit Error Rate (BER) performance severely. However, under the situation of orthogonal training and the Minimum Mean Square Error (MMSE) channel estimation, the sum ca- pacity and BER performance are consistent with those of the perfect Channel State Information (CSI) with only a performance degradation.
文摘The conventional zero-forcing vector precoding can provide the same SNR for decentralized receiving users.In order to meet different Qos requirement of multimedia applications,we modified the conventional precoding matrix by using a diagonal matrix named relative SNR control matrix so that different SNR can be got for different users＇ requirements.Then,with the new precoding matrix used,we analyze the average SNR of all MTs.We find that if a proper perturbation vector is chosen under the power constraint,the average SNR can be maximized.Finally,we indicate that the choice of perturbation vector is a K-dimensional problem which has low complexity algorithm to solve.Simulation shows that our scheme is very effective for multimedia applications.
