Abstract
The thesis proposed 5 algorithms for spatial modulation detector in polar coordinate. There are maximum likeliood detector in polar coordinate(PML), gradient detector in polar coordinate(PGML), one-stage candidate selection(OCS), two-stage candidate selection(TCS) and modified two-stage candidate selection(MTCS). We derived the maximum likelihood detector in polar coordinate(PML) and find that the complexity is lower than in rectangular coordinate. With the mathematical proof of only single local minimum in maximum likelihood distance formula and the slope of distance is monotonically increasing from the local minimum, we reinterpret the maximum likelihood solution from minimum distance to minimum slope summation. The interpretation leads to the PGML and it has lower complexity and the performance is near optimal. From the formula of slope summation, we proposed OCS algorithm by using part of the information: phase as parameter to select candidates from all the combinations of spatial and symbol index. To enhance the power of candidate selection, we proposed the TCS algorithm by adding the magnitude information which requires more complexity. With the property of convex in phase and magnitude domain, we proposed the MTCS algorithm which reduced the complexity efficiently. In the simulation result, we can see that the BER performance of sub optimal detectors in polar coordinate are near to optimal detector and the complexity are comparable to the sub optimal detector in rectangular coordinate. Among all the detectors, MTCS algorithm has the lowest complexity.