Research And Application On MIMO Detection Technology Of IEEE 802.11 Ac RF Conformance Test

As the key technology of IEEE 802.11ac, MIMO makes great contributions to enhance the data throughput of wireless network. The detection module applied in MIMO receiver will affect the performance and complexity directly. This paper mainly focuses on the MIMO detection of 802.11ac standard. Main contributions are as follows:Some basic researches on the 802.11ac are conducted in this paper, including the new features, the physical layer specifications, the frame format and the transmitter modules. MIMO and OFDM are the key techniques in 802.11ac, and the spatial multiplexing and spatial diversity are the two classes in MIMO techniques. The compromise between the two classes is also described in this paper.Some traditional MIMO detecting algorithms are studied in this thesis, including maximum likelihood(ML), sphere decoder(SD), linear, successive interference cancellation(SIC), and QRD. Both the performance and complexity of them are discussed.The sorting mechanism is introduced due to the error propagation in layered algorithms such as the ordered successive interference cancellation(OSIC) and the sorted QR decomposition(SQRD). Both of them detect the layers with large SNR firstly, and then eliminate the known interference, thereby, the accuracy of the subsequent detection is improved.The inverse matrix calculation in OSIC results in high complexity, accordingly, we propose an improved algorithm which compares the norm value in each step, combining the norm-sorted detection and OSIC. Simulation results show that the improved algorithm achieves lower complexity than OSIC and its performance is in the middle of those of the two combined algorithms. Furthermore, the complexity and performance changes along with the weighting factor, so that different factors could be selected according to various scenarios.Iteration is also introduced to obtain higher accuracy for the previous detected layers in layered MIMO detection algorithms. For example, the iterative QR decomposition(IQRD) algorithm eliminates the interference from the first layer or adjust the column vector in channel matrix by iterations to obtain better performance. ISQRD combines the iteration and SQRD, thereby giving a solution to the issues of error propagation and low diversity gain in previous detected layers. However, higher computational complexity achieves due to the process of iteration, hence an improved algorithm is proposed in this paper. The improved algorithm establishes a decision criterion to decrease the number of iterations, meanwhile solving the issues that SQRD can only get suboptimal solutions. Simulation results show that the improved algorithm performs between SQRD and iterative algorithms with reduced complexity compared to ISQRD.In addition, the research on the application of detection algorithms in RF testing system is conducted. Some algorithms about the testing items are presented briefly. Especially we build the baseband receiver for the system and focus on EVM. The main detection algorithms studied in this paper is applied to the system, and the performance with different MCS as well as the bandwidth is analyzed. Results show that the two proposed algorithms are in compliance with the requirements of RF conformance testing system.