Method for detecting spherical decode based on depth-first search

Abstract

The invention provides a method for detecting spherical decode based on depth-first search, which comprises the following steps: A, performing QR deposition on a channel matrix; B, multiplying a conjugate transpose and a received signal of a Q matrix to obtain an equalizing signal rho; C, setting an initial search radius; D, executing the depth-first search according to the initial search radius, an R matrix and the rho, and updating the search radius; E, setting an upper limit value M of the search total node number and an upper limit value Ki of the ith layer search node number; F, executing the depth-first search according to the current search radius, the R matrix and the rho, and when the search gets to the ith layer, judging whether the searched node number on the ith layer is equal to Ki or not, otherwise, executing the search on the ith layer, and if so, getting to search the i+1th layer; and G, repeating the step F until the searched total node number is equal to M or all the layers cannot be searched any more, and outputting a decoding result. The method can effectively reduce the operation complexity of the sphere decoding, and is easy to achieve through hardware.

Description

Technical field [0001] The present invention belongs to the field of wireless communication, and particularly relates to a depth-first search-based sphere decoding detection method for multiple input multiple output (MIMO) systems. The present invention can also be applied to orthogonal frequency division multiplexing (OFDM) and MIMO Detection of MIMO signals in the system. Background technique [0002] MIMO maximum likelihood (ML) detection can make the system obtain the best bit error rate performance, but the ergodic search is often difficult to implement in real time in actual systems due to its non-deterministic polynomial (NP) computational complexity. It cannot be achieved, and low-complexity MIMO-ML and signal detection algorithms close to ML have always been problems to be solved by MIMO systems. [0003] Based on the research of Pohst et al., Viterbo et al. proposed a detection algorithm called sphere decoding for source signals with grid-like constellations. The sphere...

AI Technical Summary

Problems solved by technology

Hassibi and Vikalo show that sphere decoding has polynomial computational complexity

[0005] Nevertheless, there are still many deficiencies in the sphere decoding algorithm, which affect the practical application of the algorithm

First of all, the selection of initial radius is very important for spherical decoding. Too large initial radius will lead to excessive computational complexity, while too small initial radius will cause no emission signal mapping point in the sphere and the search will fail.

Then, the number of grid points in sphere decoding is a random variable, which is affected by channel conditions, noise and initial radius, so it is difficult to quantitatively analyze the detection complexity and delay. As pointed out by Jalden et al., the operation of sphere decoding is complex The degree is still NP in the worst case, which is very unfavorable in real-time high-speed applications; in addition, the uncertainty also makes the hardware implementation of sphere decoding difficult

Images