Mimo-ofdm Systems Signal Detection Techniques And Vlsi Implementations

MIMO(Multiple-Input-Multiple-Output) wireless communication system is capable of providing much more system capacity compared with traditional single-input-single-output system,which has become the hot spot of communication system.MIMO technology is widely used in communication field such as 3G,LTE or 4G.For example,4×4 64-QAM MIMO system is used in IEEE802.16e which is able to achieve 70Mbps throughput.The throughput of IEEE802.11n is nearly 500Mbps with the same system condition.However,with the increasing in antenna quantity as well as the modulation complexity,the hardware complexity of MIMO system is also increased simultaneously.As a matter of fact,how to design a signal detector in MIMO system with low hardware complexity and high speed data transfer capability has become the topic in wireless communication field.In addition,with the spread of MIMO technology in mobile electronic devices area where SNR is always changing, it is much more advantageous to the quality of communication that the architecture with reconfigurable multiple modulation is applied in a single system.It means that the MIMO signal detectors need to have the ability of demodulating different modulations with the same hardware structure.In order to meet the requirement of high speed,low complexity and reconfigurable structure that are discussed above,two MIMO signal detectors:Channel-Aided Pruning-Ahead Detector and Pre-Pruning Folding K-Best Detector are presented in this paper.Current MIMO-OFDM signal detector algorithms are analyzed firstly.Both the algorithms presented in this paper are based on K-Best algorithm,since it performs a breadth-first searching process in signal detection that is able to achieve a sub-ML detection performance and has the characteristic of parallel calculation without iteration,which makes it be suitable to the high speed MIMO communication system and meet the requirement of wireless communication applications in next generation.When quantity of antenna arrays is increased,the parameter K in K-Best algorithm, which represents the number of signal nodes to be stored and outputted in each detection layer,must be increased too.To make it worse,K-Best algorithm expands and calculates all child nodes of K stored nodes,which make it unlikely to avoid the defect that the computational complexity will linearly increase with the complexity of modulation methods.For example,huge hardware resources will be consumed when implementing a K-Best signal detector in a 4×4 64QAM MIMO system in some protocol such as IEEES02.11n.To overcome this drawback,the two algorithm presented in this paper is aiming to decrease the computational complexity of signal detection.The Channel-Aided Pruning-Ahead Detector deletes the child nodes in advance with the aid of channel information and characteristics while the Pre-Pruning Folding K-Best Detector prunes the child nodes and calculates the remaining ones in folding mode.Both of the detectors have decreased the computational complexity dramatically without the degradation of detection performance.The hardware implementations of CAPA detector and PPFKB detector are presented in this paper.Controlled by the input signal,CAPA detector and PPFKB detector are able to process signal detection in QPSK/16-QAM modulation and QPSK/16-QAM/64-QAM modulation respectively.The parallel and pipelined structures are implemented so as to improve the detection speed.Module multiplexing, timing matching is used to reduce the consumption of hardware resource.CAPA MIMO signal detector is implemented on FPGA platform of Xilinx Virtex-4 FPGA(XC4VLX200).It has achieved a 2Gbps throughput in 16-QAM modulation, which has reached a high level result among the detectors with the similar specifications.PPFKB MIMO signal detector is implemented in ASIC flow with SMIC 0.18μm library.It has achieved a 1Gbps throughput in QPSK/16-QAM/64-QAM modulation,which has also obtained a very high performance.With the design of algorithms and optimization of hardware structures, two MIMO signal detectors presented in the paper have the great advantage in performance.