King Abdullah University of Science and Technology

Research Center on Scientific and Technical Information

David Elliot Keyes

Mohamed-Slim Alouini

Hatem Ltaief

Zouheir Rezki

Mohamed Amine Arfaoui

Adel Dabah

A detection method is presented that combines two non-linear approaches, the Sphere Decoder (SD) and the K-best algorithms, to identify symbols transmitted over a communication channel in a multiple input-multiple output (MIMO) communication system. A parallel implementation on a x86 computer system is presented, which reduces the number of evaluations by branching over plural symbols and diversifies the search process, which may rapidly reduce the radius and thus, the time complexity of the computer system.

Massive multiple-input multiple-output system and method

A method of detecting a symbol transmitted over a communication channel in a multiple input-multiple output communication system. The method includes receiving a plurality of symbols transmitted over a communication channel of a multiple input-multiple output communication system. A sphere radius is initialized based on attributes of the communication channel. A first matrix of possible transmitted signals is defined as well as a second matrix of received symbols. The matrix of possible transmitted signals is searched using a breadth-first search (BFS). Each level of the search tree is analyzed utilizing matrix multiplication to determine selected symbols satisfying the initialized sphere radius. A maximum likelihood solution is of the transmitted symbols is derived based on the selected symbols.

Efficient sphere detector algorithm for large antenna communication systems using graphic processor unit (GPU) hardware accelerators

A method of detecting u symbol transmitted over a communication channel in a multiple input-multiple output communication system. The method includes receiving a plurality of symbols transmitted over a communication channel of a multiple input-multiple output communication system. A sphere radius is initialized based on attributes of the communication channel. A first matrix of possible transmitted signals is defined as well as a second matrix of received symbols. The matrix of possible transmitted signals is searched using a breadth-first search (BFS). Each level of the search tree is analyzed utilizing matrix multiplication to determine selected symbols satisfying the initialized sphere radius. A maximum likelihood solution is of the transmitted symbols is derived based on the selected symbols.

Growing community of inventors

Thuwal, Saudi Arabia

David Elliot Keyes