Near maximum likelihood detection algorithm based on 1-flip local search over uniformly distributed codes

The maximum likelihood (ML) detection is the process to find the nearest lattice point to a given one in an N-dimensional search space. The ML problem is well known to be NP-hard. In this paper, we propose a near-maximum likelihood detection algorithm based on an intensification strategy over an initially efficient and uniformly distributed subset. This subset is given by a diversification step based on powerful uniformly distributed codes. The proposed algorithm has three characteristics that make it attractive for several practical wireless communication systems. First, the simulated bit error rate performance shows that this algorithm provides a good approximation to the ML detector. Second, it has a constant polynomial-time computational complexity. Finally, the inherent parallel structure of this algorithm leads to a suitable hardware implementation.