Over-the-Air DPD and Reciprocity Calibration in Massive MIMO and Beyond

Ashkan Sheikhi,Ove Edfors,Juan Vidal Alegría
2024-11-07
Abstract:In this paper we study an over-the-air (OTA) approach for digital pre-distortion (DPD) and reciprocity calibration in massive multiple-input-multiple-output systems. In particular, we consider a memory-less non-linearity model for the base station (BS) transmitters and propose a methodology to linearize the transmitters and perform the calibration by using mutual coupling OTA measurements between BS antennas. We show that by only using the OTA-based data, we can linearize the transmitters and design the calibration to compensate for both the non-linearity and non-reciprocity of BS transceivers effectively. This allows to alleviate the requirement to have dedicated hardware modules for transceiver characterization. Moreover, exploiting the results of the DPD linearization step, our calibration method may be formulated in terms of closed-form transformations, achieving a significant complexity reduction over state-of-the-art methods, which usually rely on costly iterative computations. Simulation results showcase the potential of our approach in terms of the calibration matrix estimation error and downlink data-rates when applying zero-forcing precoding after using our OTA-based DPD and calibration method.
Signal Processing,Information Theory
What problem does this paper attempt to address?
This paper attempts to solve the problems of non - linearity and reciprocity calibration of the base station (BS) transmitter in the massive multiple - input multiple - output (Massive MIMO) system. Specifically: 1. **Non - linearity compensation**: The non - linear response of the base station transmitter will cause signal distortion and affect the performance of the system. Traditional digital pre - distortion (DPD) methods usually rely on special hardware modules to measure the input - output characteristics of the amplifier and design an inverse function to eliminate non - linear effects. However, these methods require additional hardware resources, increasing the complexity and cost of the system. 2. **Reciprocity calibration**: In actual deployment, the hardware differences between the base station's transmitting and receiving links will break the ideal channel reciprocity assumption, thus affecting the accuracy of downlink precoding. To solve this problem, reciprocity calibration is usually required. Most of the existing reciprocity calibration methods assume that the base station's transmitting link is linear, which is not always true in Massive MIMO systems, especially when using low - cost, non - linear components. To solve the above problems, this paper proposes an over - the - air (OTA) - based method to simultaneously achieve DPD and reciprocity calibration by using the mutual coupling measurement data between the base station antennas. The main advantages of this method include: - **No special hardware required**: Through OTA measurement, non - linearity compensation and reciprocity calibration can be achieved without additional hardware. - **Simplified calibration process**: Compared with existing methods, the proposed OTA method can significantly reduce the computational complexity and avoid complex iterative algorithms. - **Applicable to non - linear systems**: This method is not only applicable to linear systems, but can also effectively handle the non - linear characteristics of the base station's transmitting link. In conclusion, this paper proposes an innovative OTA method that can effectively perform non - linearity compensation and reciprocity calibration in Massive MIMO systems, thereby improving the performance of the system and reducing costs.