The BEAST for Maximum-Likelihood Detection in Non-Coherent MIMO Wireless Systems

Next generation wireless systems have to be able to efficiently deal with fast fading environments in order to achieve high spectral efficiency. Using multiple-input multiple-output (MIMO) systems and exploiting receive diversity, the spectral efficiency can be greatly increased. Commonly, the channel is estimated via training symbols, before data detection is carried out based on the obtained channel estimate. While this significantly simplifies the process of data detection, it leads in general to suboptimal results. A better approach is to carry out joint channel estimation and data detection; we turn our attention to joint maximum-likelihood (ML) detection which is the optimal strategy. In this paper, the BEAST - Bidirectional Efficient Algorithm for Searching code Trees - is proposed as an alternative algorithm for joint ML channel estimation and data detection and its complexity is compared with recently published algorithms in the literature.