Sphere Decoding for Multiple Antenna Systems


In multiple-input multiple-output (MIMO) communication systems the deployment of several antennas at the transmitter and receiver introduces additional degrees of freedom (spatial dimension) into a wireless communication system. These additional degress of freedem allow to substantially increase the data rate (bit/sec) or to improve the transmission quality (bit error rate) in a wireless link [1]. There are two basic space-time processing methods which make use of these degrees of freedom, namely space-time coding to improve link reliability and spatial multiplexing to increase spectral efficiency. With spatial multiplexing it is possible to enhance the data rate without additional effort in bandwidth or power by transmitting parallel substreams simultaneously over spatial subchannels [1].

The receiver that is able to optimally recover the interfering data streams is the Maximum Likelihood (ML) receiver. Unfortunately, the complexity of the ML receiver is exponential in the number of transmit antennas and length of channel memory and is therefore often not applicable in practice. However, the ML detection problem can be solved via the Sphere decoding algorithm [2] with expected complexity that is only polynomial. The idea of Sphere Decoding is to search only for the nearest lattice point (represents a hypothesis) within a hypersphere of radius R centered at the received signal instead of an exhaustive search over all lattice points.

In this project you will study and implement the Sphere Decoding algorithm for flat-fading and dispersive MIMO channels. Your task will be to compare the performance-complexity tradeoffs of the Sphere Decoding algorithm for different Space-Time code designs.

[1] D. Gesbert and J. Akhtar, "Breaking the barriers of Shannon's capacity: An overview of MIMO wireless systems", Telektronikk Telenor Journal, Jan. 2002.
[2] B. Hassibi and H. Vikalo, "On the Sphere Decoding Algorithm, Part1 & Part2".

Subject area MIMO communication systems, Space-Time receiver
Type of work 50% Theory, 50% Software
Student Mirsad Cekic, Adjan Kretz
Supervisor Dr. Ingmar Hammerström, Dr. Boris Rankov
Professor Prof. Dr. Armin Wittneben