Orderly Distributed Cellular Structure And Distributed Multi-antenna Techniques

Distributed cellular architectures and distributed MIMO techniques become the "Hotpoint" in the wireless research field during resent years. Also, they will play an important role in the next generation wireless communication systems such as the 3G LTE. In this work, a novel distributed cellular architecture was proposed and its capacity was analyzed. Furthermore, we did a deep research work on the corresponding distributed MIMO techniques.Firstly, a novel distributed cellular architectures was proposed and in different parts of the cell, different transmit power control schema was adopted. And then, the capacity of this novel architecture was investigated in the single-user and multi-user systems compared with the traditional cell and the so-called full distributed architecture, which was proposed by other researchers. Through numerical simulation, the results showed that the novel architecture had the best performance. And then, we analyzed and proved that it is the distributed antenna selection make this novel architecture has the best performance.Secondly, in this work, cyclic delay diversity (CDD) technique was researched. In the block fading channel, the relationship between effective channel power and the upper bound of conditional pair-wise error probability (PEP) was derived to show the performance gain of CDD, and an adaptive transmit technique was proposed. Based on the channel estimation at receiver, the transmitter is informed whether or not to use CDD by 1-bit feedback, which brings the diversity of two kinds of channels. The numerical results showed that with this adaptive technique, the BER performance was improved. With the contribution from other research, CDD performance was analyzed in distributed cell, and the power allocation was considered to optimize the BER performance.In this work, using Cyclical Delay Diversity (CDD), which making the channel more frequency-selective by cyclically delaying the data copy in multiple transmit antennas, we proposed a frequency domain random beamforming method to provides more opportunity for the users in poor condition. The beamforming vectors distribute uniformly in the linear space that is spanned by the users' frequency-domain channel vectors. Furthermore, this beamforming method has a simple structure to produce, and is easily supported on both side of the transceiver. The numerical simulation results showed that using sub-carrier based scheduler this beamforming method enhanced the system throughput performance and improved the fairness among users.Thirdly, it is well known that the distributed antenna connects with independent amplifier so that every antenna has a power limit is more practical. The maximum sum capacity of zero-forcing (ZF) precoding under per-antenna power constraints (PAPC) can be solved by using a barrier based interior-point method. Rather than looking for the optimal power allocated to users, namely the user power optimum (UPO), in the downlink, we try to find the optimal power for the transmit antennas alternatively and we call this transformed problem the antenna power optimum (APO). Through the theoretic analysis and numerical simulation, it has been proved that APO rather than UPO can achieve the exact capacity of ZF precoding under PAPC. This result implies that in order to achieve a correct capacity curve of the ZF precoding, it is APO rather than UPO should be adopted when solving the problem by using a barrier function based interior-point method.