MIMO GPU Detector

Multiple input multiple output (MIMO) greatly improves the throughput of communication systems by using multiple antennas for the transmitter and the receiver. Detectors based on computationally intensive searches are needed to derive maximum performance from MIMO systems. To solve this problem, custom designs on ASIC or FPGA are often used in receivers. The challenge is further complicated by the fact that the receiver needs to handle various configuration of typical industry standards. The graphics processor unit (GPU) is a low-cost parallel programmable coprocessor that provides extremely high computational throughput for many signal processing applications.

Large MIMO (Multiple Input Multiple Output) antennas use multiple antennas and multi - user MIMO (MU - MIMO) to increase sector throughput and capacity density. Each antenna is individually controlled and can be embedded in the radio transceiver assembly. Nokia claims that the capacity of the 64-Tx / 64-Rx antenna system has increased five-fold. The term "large-scale MIMO" was first created by Dr. Thomas L. Marzetta, researcher at Nokia Bell Labs in 2010, and was launched on 4 G network such as Softbank in Japan.

In recent years, antenna technology has achieved incredible improvements, namely Massive MIMO and MU-MIMO. Large-scale MIMO (Multiple Input Multiple Output) technology refers to the idea of ​​equipping a radio transceiver with multiple antennas and has been shown to improve the spectral efficiency of radio links. Massive MIMO essentially extends the MU - MIMO (Multi - User MIMO) concept already popular in LTE networks by increasing the number of antennas far more than the number of simultaneous users.

Large-scale MIMO with beam steering: MIMO is used in many applications from LTE to Wi-Fi, but the number of antennas is quite limited. The use of microwave frequencies opens up the possibility of using dozens of antennas on a single device, which is realistic due to antenna size and wavelength spacing. This allows you to manipulate the beam to improve performance. High Density Network: Reducing the size of the cell allows for more efficient use of available spectra. It is necessary to ensure that small cells in the macro network and small cells arranged as femto cells can operate satisfactorily. Adding a large number of cells to the network has a big challenge and technology for that has been developed.