Research On Beamforming Technology For Wireless MIMO Systems

In a multiple antennas system, beamforming is a versatile and powerful approach tosuppress interference, which makes it being one of the key physical layer technologies of thenext generation broadband wireless MIMO communication system, as well as gaining theoverwhelming interst of many studies. This paper focuses on the beamforming technology forwireless MIMO systems. On the basis of taking MIMO system structures, channel characteristicsand practical applications into account, the beamforming technology is deeply analyzed in thispaper for MU-MIMO downlink systems, cellular networks MIMO downlink systems and relaynetwork systems, which aims at providing theoretical support and solutions for beamformingproblems of MIMO systems wireless transmitting. To be specific, the main work of this papercan be summarized into the following aspects:(1) Specific to joint design problem of transceiver beamforming and power allocation in theMU-MIMO downlink systems, an iteration strategy based on fix-point iteration is proposed fortwo different design criteria: minimizing the total transmitting power and maximizing theminimum SINR. By utilizing the fixed point iteration, the proposed scheme obviates complexeigenvector problem, which reduces the computation complexity. And then, the fixed pointiteration is studied further by making use of the uplink-downlink duality, which demonstrates therelationship between fix-point iteration objects and uplink transmitting power, and illustrates thevariables coupling relationship and the feasibility of the problem. Based on the further analyzing,an improved algorithm is proposed, which has greatly lower complexity due to the transform ofthe complexly transmitting beamforming into receiving beamforming. Numerical simulationdemonstrates that the proposed two schemes have better convergence behavior, and greatlyreduce calculating accuracy while getting the same performance with the existing algorithms,which means that, with the same calculating accuracy, better performance can be obtained fromour algorithms. Thus, the proposed two schemes have more practical value with respected tocomputation complexity and computational accuracy.(2) Robust beamforming against imperfect CSI in MU-MIMO downlink systems is thefocus of many studies. However, to the best of our knowledge, when the imperfect CSI isformulated as NBE model, little attention has been paid to the uplink-downlink duality, which iswell known to provide flexible and efficient solution for beamforming under perfect CSI.Therefore, on the basis of finding out the conservative approximation of Worst-Case SINR, thispaper establishes the Worst-Case SINR uplink-downlink duality. Using this uplink-downlinkduality, a scheme based on SDP program is proposed for power minimization problem. With theaids of this uplink-downlink duality, the scheme always transforms complex transmittingbeamforming problems into receiving beamforming, and its SDP programming has only Lvariables，which greatly reduce the computational complexity of the algorithm. Besides, basedon eigenvector problem and linear equations problem, this paper also proposes interactiveschemes for SINR balancing problem and power minimization problem, respectively. Theschemes use eigenvector problem and linear equations problem, respectively, instead of SDP programming, which avoids interior point iteration of the convex optimization. Thus, thecomputational complexity of the algorithm is further reduced. Numerical simulation resultsdemonstrate the robustness of the above-mentioned algorithms against imperfect CSI.(3) Specific to cellular network downlink beamforming design, firstly, this papergeneralizes the uplink-downlink duality of the MU-MIMO systems to cellular network MIMOsystems. And then by using the duality, a design strategy based on interactive iteration isproposed for cellular network power minimization problem and SINR balancing problem. Incellular network mobile systems with the presence of micro-scale fading and per-cell powerconstraints, this paper establishes the uplink-downlink duality, based on the seconde orderstatistics of CSI, for total transmit power minimization problem with per-cell power constraints.With the aids of this duality, an effective inner-outer iterative algorithm is proposed forbeamforming design in cellular network MIMO downlink systems. The simulation results showthat, without the per-cell power constraints, the beamforming design may exceed the maximumtransmitting power of base station, which is unrealizable.(4) Specific to one-way relay and two-way relay of relay network systems, this paperstudies the joint design problems of relay network beamforming and power allocation under thecriterions of SNR maximization and SNR balance, respectively. For one-way relay beamformingproblems, by using the inverse conclusion of the SNR maximization problem and the powerminimization problem, an algorithm is proposed based on bisection searching. And for thetwo-way relay beamforming problem, based on the equality of the SNR balance problem and theone-way relay SNR maximization problem, a scheme is proposed by constructing virtualone-way relay system. Besides, by using the equality, this paper also derives the relationshipbetween one-way relay and two-way relay, which demonstrates that the maximum transmissionrate of two-way relay is higher than that of one-way relay. Numerical simulation results show thecorrectness of the theoretical derivation.