Iterative Tree Search Precoding for Multiuser MIMO
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-user multi-input multi-output (MIMO) downlink communication systems face challenges in efficiently eliminating interference between users, leading to high complexity and suboptimal performance due to the need for advanced precoding techniques and channel state information management.
Innovation Solution
An iterative tree search-based precoding method is introduced, which limits the search candidate domain based on channel state information, calculates a reference value, and applies a distortion value for modulo operation, reducing complexity and improving performance by grafting the technique onto the receiver side.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Sphere Encoding algorithm is used to achieve optimal performance, then performance is improved, but complexity increases
Solution Approach 1:
The patent segments the precoding process into two distinct stages: (1) interference elimination using linear precoding matrix, and (2) constellation expansion using modulo operation. This segmentation allows each stage to be optimized independently, achieving near-optimal performance with reduced overall complexity compared to the monolithic Sphere Encoding algorithm.
Solution Approach 2:
The patent applies preliminary interference elimination using linear precoding before the signal is transmitted. By pre-processing the signal to eliminate interference between users at the transmitter side, the system achieves better performance without requiring complex detection algorithms at the receiver, thus reducing overall system complexity.
2Device complexity
If linear precoding technique is used to eliminate interference, then device complexity is reduced, but performance degrades due to power loss
Solution Approach 1:
The patent changes the signal representation parameters by expanding the constellation to infinity through modulo operation. This parameter change allows the system to overcome the power loss limitation of linear precoding by mapping signals to an extended constellation space, thereby improving performance while maintaining low complexity.
Solution Approach 2:
The patent introduces an intermediary distortion value that is added to the precoded signal before modulo operation. This intermediary element acts as a bridge between the linear precoding output and the final transmitted signal, enabling the system to correct power loss effects and improve performance without increasing complexity.
3Reliability
If Tomlinson-Harashima Precoding is used to restore original information, then performance is improved compared to linear technique, but optimal performance is not achieved
Solution Approach 1:
The patent inverts the traditional approach by applying modulo operation at the transmitter side rather than at the receiver side. This inversion allows the system to achieve better performance by pre-distorting the signal in a controlled manner, making the reception process simpler while achieving near-optimal performance.
Solution Approach 2:
The patent creates a universal precoding framework that combines interference elimination and constellation expansion in a single process. This multi-functional approach allows the system to achieve both interference cancellation and power loss compensation simultaneously, providing near-optimal performance with a unified low-complexity structure.
Data Source
AI summary
An iterative tree search-based preceding method for a multi-user Multi-Input Multi-Output (MIMO) communication system includes determining a reference value of a cumulative branch metric of a candidate symbol, eliminating candidates having values that exceed the determined reference value of the cumulative branch metric of the candidate symbol, and registering values, which do not exceed the determined reference value of the cumulative branch metric of the candidate symbol, as entries, and selecting the least value from the values registered as the entries. Thereby, the method has low complexity and similar performance compared to an existing Sphere Encoding (SE) technique.