Transmitting power minimization transmission method based on reconfigurable intelligent surface

Abstract

The invention discloses a transmitting power minimization transmission method based on a reconfigurable intelligent surface, and the method comprises the steps: enabling a signal transmitted by a base station to reach a user side through the reflection of the reconfigurable intelligent surface, and enabling the reconfigurable intelligent surface to change the phase of a signal incident on the reconfigurable intelligent surface, so as to enhance the receiving signal of the user side. Therefore, the effect of reducing the transmitting power of the base station end on the premise of ensuring the service quality is achieved. According to the method, a base station and a reconfigurable intelligent surface utilize known channel state information, based on a transmitting power minimization principle, a non-shared subcarrier distribution strategy among users is considered, and on the premise that user performance constraints are met, the user performance is improved. A transmitting end precoding matrix and a reconfigurable intelligent surface reflection phase matrix are jointly designed through a maximum ratio transmission precoding scheme and a deep reinforcement learning algorithm. According to the method, the convergence speed is high, and compared with a traditional numerical method, the transmitting power is effectively reduced through low calculation complexity and time delay.

Description

technical field [0001] The invention belongs to the technical field of adaptive transmission of a multi-user multiple-input single-output orthogonal frequency division multiplexing (multiple-input single-output orthogonal frequency division multiplexing, MISO-OFDM) downlink system assisted by a reconfigurable intelligent surface, in particular to a Transmit power minimization transmission method based on reconfigurable smart surface. Background technique [0002] Benefiting from the rapid development of metamaterials, reconfigurable intelligent surface (RIS) is regarded as one of the key technologies for achieving high spectral efficiency at low cost and low energy consumption in the new generation of wireless communications. RIS is a software-controlled planar array metasurface composed of a large number of passive and reconfigurable reflective elements. By dynamically adjusting the amplitude and / or phase of the reflected signal in real time, the reflected signal energy is ...

AI Technical Summary

Problems solved by technology

[0003] However, the number of RIS reflective elements deployed in practice generally varies from thousands to tens of thousands and there are constant modulus constraints at RIS, so how to improve communication performance through joint design in RIS-assisted wireless communication systems is a multi-task problem

Traditional design methods include successive convex approximation and positive semi-definite relaxation algorithms, etc. These traditional algorithms can show better performance when the number of RIS reflective elements is small, but with the increase of the number of reflective elements, the system complexity is too high, and It will introduce a large communication processing delay and cannot meet the needs of real-time communication

If these traditional numerical algorithms are applied to broadband systems, the design complexity will be further increased. This is mainly due to the inconsistency of the subcarrier channels in the broadband system, and the optimal design of a single subcarrier channel will no longer be optimal for other subcarriers. , which requires a joint design algorithm with low time cost and high performance to further solve these problems

[0004] Existing research on low-complexity algorithms for reflection matrix design at RIS is mainly based on the design of reflection patterns at RIS or the use of deep learning methods. The design of reflection patterns at RIS needs to be based on multiple channel experiments to obtain the best pattern. The learning method has high requirements on the amount of training data and is sensitive to channel fading changes

These methods sacrifice a certain system performance in exchange for a reduction in time complexity, which cannot achieve a win-win goal and is difficult to apply in practice.

Images