Abstract :
The switched a posteriori probability (s.a.p.) diversity combiner has equivalent performance to maximum a posteriori probability (m.a.p.) combiner for dual diversity when binary transmission is used. Moreover, it is much simpler to implement since it is similar in principle to the conventional switched diversity combiner. The difference is that instead of switching to the branch with the highest signal-to-noise ratio, the switching is done between the branches on a bit-by-bit basis and to the branch which is most likely to be correct about what was sent. Some examples of impulse noise are used to show the difference between the m.a.p. (or s.a.p.) methods and the conventional methods of maximal ratio combining, selector branch combining, and no diversity, in terms of average error probability. The examples also illustrate the different processing that is done on the signals by the different methods. One possible method of implementing the s.a.p. technique is proposed for the case of 4-phase p.s.k. with differential encoding.
