Millimeter Wave Channel Estimation Using Compressive Sensing

The patent changes the parameter of angle sampling from regular intervals to non-uniform intervals based on compressive sensing theory. By transforming the uniform sampling approach into a non-uniform sampling approach with specific mathematical properties, the system achieves accurate channel estimation with fewer measurements. The key parameter change is in the sampling strategy itself, using optimized angle intervals that satisfy compressive sensing conditions rather than conventional uniform spacing.

The patent replaces the traditional exhaustive search mechanical process with a mathematical transformation approach. Instead of physically performing numerous measurements at regular intervals and processing them through exhaustive search algorithms, the system uses compressive sensing mathematical models to reconstruct channel information from fewer, strategically selected measurements. This substitution of mathematical theory for mechanical measurement processes significantly reduces time consumption.

The patent changes the fundamental parameter of how beamforming vectors are selected and structured. Instead of using a large number of vectors at regular angle intervals, the system employs a reduced set of vectors with specifically optimized non-uniform angle spacing that satisfies compressive sensing conditions. This parameter change in the angular distribution and selection criteria of beamforming vectors achieves high resolution without proportionally increasing calculation complexity.

The patent applies partial action by using only the minimum necessary number of beamforming vectors required for accurate channel estimation, rather than exhaustively testing all possible angle intervals. The compressive sensing approach identifies and utilizes only the critical measurements needed, performing partial measurements and calculations that are sufficient to reconstruct the full channel information, thereby reducing overall computational burden.