More Uplink Pilot Pollution In Village Tdd System Analysis

In the CoMP existing study, most assume that the CSI is known or correctly estimated by the receiving end, but in the real system, it is impossible. In the existing MIMO-OFDM single base station cellular systems (3GPP R8standard), the base station using different resource partitioning allows edge users to occupy different subcarriers, that means different frequencies to achieve networking to avoid the pilot pollution problem. But the CoMP system’s goal is to achieve the same frequency network, different cell edge users using the same frequency, which will cause the pilot signals interfere with each other.By using the channel reciprocity in TDD system, base station estimate uplink channel to get downlink CSI. So it is necessary to research the uplink non-orthogonal training sequence for channel estimation error and system capacity influence.This paper shows that the SRS signal in different cells may appear non-orthogonal in three cases.we also simulate the estimation error when using orthogonal and non-orthogonal training sequence in neighboring cells. The results show that the estimation error is large when use non-orthogonal training sequence, and more users will lead to stronger interference and also more serious CSI distortion. Uplink channel estimation errors affect the system capacity through the calculation of the precoding, The simulation results show that the channel estimation error make the system rate reduced significantly, and prevented by increasing the uplink transmission power to increase capacity;in addition, the large-scale fading of interactive channel has a great influence on the channel estimation error and capacity.In view of the non-orthogonal training sequences in CoMP capacity decline, this paper put forward two kinds of orthogonal SRS signal scheme. First scheme based on the SRS configuration index, it is easy to realize but less number of users will have certain resources waste, the second scheme make all cells use those SRS sequences which is based on the same base sequence; using these sequences, Inter-cell edge users keep orthogonal in CDM; the spectrum efficiency is higher than the first one, but it is more complex. Finally we introduce a multi-cell MMSE-based precoding algorithm which take channel estimation error into consideration, The simulation results show that this algorithm is superior to ZF algorithm and SLNR algorithm, and stronger interference will get more obvious advantages.