Description

This pre-processing allows for employing a tree-based search algorithm at the second stage that finds the optimal symbol which has the minimum distance to the received symbol. In  this  chapter,  we  present  an  efficient  FPGA  prototype  of  the  FSD  that  implements  an efficient QR decomposition of the channel matrix for a MIMO system using CORDIC array techniques  followed  by  the  FSD  tree-based  search  algorithm.  This  prototype  provides  an improved  hardware  implementation  in  terms of  area  and  throughput  compared  to  up-to-date  published  results  in  the  literature.  The FPGA  implementation  is  incorporated  with  a MATLAB  simulation  model  of  an  OFDM  based  MIMO  system  to  validate  the  hardware design. As a practical application, we employed the FSD in the receiver of the IEEE802.16e WiMAX MIMO system. The IEEE 802.16e standard supports three types of MIMO space time codes(STC), referred to as matrix A, B, and C. These  STC  codes  achieve  different  levels  of throughput  and  diversity  depending  on  the  quality  of  the  MIMO  channels.  We  employed the  FSD  as  a  STC  detector  to  decode  the  different  STC  schemes.  In achieving  that,  the received signal is adaptively pre-processed according to the STC type prior feeding it to the FSD.  At  the  end  of  the  chapter,  we  present  the  performance  of  a  two  by  two  STC-WiMAX system with the FSD in terms of bit error rate and throughput. The results indicate that our FPGA  implementation  achieves  throughput  values  that  satisfy  the  requirements  of  the WiMAX standard.