Precoding And Related Techniques Research In Multi-Antenna Mobile Communication Systems

Precoding technology, accompanied by detection cr equalization, under the assumption of channel state information is known at transmitter, acts as the pre-processing technique at transmitter side, which can mitigate the interference caused by the multi-access interference owing to multi user or multi-antenna and the inter symbol interference caused by frequency selective fading owing to multi path effect in wireless channel, accordingly, which reduced the complexity at receiver side. And for downlink channel, user equipment(UE) act as the receive terminal, owing to its limited size, power consumption and prices etc, parallelly, the equalization algorithm at UE can be arranged at Base Station(BS) side as pre-process technique, which is very important especially for reducing the complexity at UE.Different from point to piont communication mode, in relay systems, when relay node added, the whole communication process is divided into2phases, the first phase is the communication between source and relay node, and the second phase is the communication between relay and destination node, if there exists direct path between source and destination node, the relay communication systems will have cooperrative diversity gain, precoding and equalization joint optimization algorithms are deeply researched in this dissertation when "with" or "without" direct path in relay systems, the content includes:According to there exists direct path or not in three node relay systems, in chapter3, the precoding related algorithms are given in relay systems under the ideal channel state information when communication nodes equipped with multi antennas. In this part, assumed the channel estimation is perfect at receiver, moreover, there is no error appeared in the second phase’s CSI feedback from destination to relay node, as well as the first phase’s CSI feedback from relay to source node, therefore, this is the ideal case in fact, and two closed form resolutions are given in this chapter, the first one is based on MMSE criterion to the estimated signal at destination node, the lower bound performance is illutrated when the constraint condition is relaxed, even so, the performance is still better than the other linear resolution based on ZF/MMSE criterion because the joint optimization between the transfer matrix at relay node and scalar factor at destination node, the second method is based on the solvement of matrix equation with type "AX+XB=C", the optimization solution is given and compared with other related solutions, the outage probability and ergodic capacity are all impoved obviously to some extent.When there exists direct path between source and relay node, two solutions are given in the following parts in chapter3, the first one is based on the joint equalization to the signal received at destimation node of the two paths, that is the direct path together with "source-relay-destination" path, also using the iterative solution,through the joint maximum rate combination(MRC) and get the optimal performance, the second solution is as following:the equalization solution is applied to each path first, then adjust the weight factor to these two pathes, the combining method is used at last, which is different from the first method, only scalar weight is adjusted here, owing to the joint optimization to the two pathes, the equalization matrix in these two phases is contained in each other, that is not only phase but also amplitude scalar weight are adjusted, the performance is compared between these two solutions, and the cooperative diversity gain is also given between the two cases ("with" or "without" direct path in the three nodes relay systems).In chapter4, the robust precoding algorithm is given under non-ideal CSI in relay systems. Firstly, the non-ideal channel modeling method is introduced, and then the robust joint precoding and equalization optimization algorithm is supplied under the case without direct path between source and destination node, and detailedly the difference is analysed between robust and non-robust solution; take the case with direct path into account between source and destination node, it is divided into two sub-cases, the first one considers the joint optimization between source precoding and destination equalization, which can be ascribed as Schur-Concave optimization problem, and the closed resolution is given in this part; the second sub-case is related to the joint optimization between relay precoding and destination equalization, then upper and lower bound is shown, adopt the inner-point optimization method, and the resolution is given based on Matlab optimization toolbox.In chapter5, the general precoding (such as smart antenna beamforming) algorithm in cognitive radio network is consided, and this chapter is divided into two problems. Firstly, beamforming and power control joint optimization is discussed in cognitive network, and according to the different priority between primary and cognitive users, two weight generation methods are supplied, in order to ensure the performance of primary users, the weight factor is set as1, and to cognitive user, the adaptive weight factor is adopted in order to reduce the access signal-to-leakage and noise ratio(SLNR) threshold, from this scheme, the reference SLNR threshold is supplied to cognitive users. In the second part, the beamforming designing algorithm is given under non-ideal CSI cognitive radio networks, the base station is configured with multi-antenna here, while for primary and cognitive users, there is only one antenna configured, and briefly, this case only considered one primary user and one cognitive user, and this problem is ascribed as fraction optimization problem, through proper decomposition, it can be regard as the numerator and denominator seperate optimization problem. And finally it is can be ascribed to convex optimization problem, and we can get the resolution based on Sedumi (the convex optimization toolbox), this solution is better than other solution in related references, as can be seen clearly through the comparative simulation analysis.