Title :
Multiple-symbol differential detection of π/4-DQPSK in land mobile satellite communication channels
Author :
Xiong, F. ; Wu, D.
Author_Institution :
Dept. of Electr. Eng. & Comput. Eng., Cleveland State Univ., OH, USA
fDate :
6/1/2000 12:00:00 AM
Abstract :
Multiple-symbol detection of π/4-DQPSK in land mobile satellite communications (modelled as correlated Rician fading) is studied. Metrics for maximum likelihood detection are derived. For the no channel-information case. The generation of the metric can be implemented as a multiple-symbol differential detector. For the ideal channel-information case, however, the generation of the metric cannot be implemented as a multiple-symbol differential detector. The BER performances of this scheme are evaluated through the union bounds of pairwise bit error probabilities and Monte Carlo simulation for both cases. It has been found that in light or medium fading Rician channels, about 1.0 dB gain for Eb/N0 at a BER=104 can be achieved by using four-symbol detection instead of conventional two-symbol detection. The gain is greater than 1 dB when the BER is lower. It is also demonstrated that multiple-symbol differential detection actually increases the BER for severe fading Rician channels (K<10.5 dB) or Rayleigh channels if no channel information is available. However, with ideal channel information, the multiple-symbol detection can reduce the BER at high Eb/N0. The influence of different fading rates on the BER performance is also evaluated, showing a lower BER at a lower fading rate, as expected
Keywords :
Monte Carlo methods; Rayleigh channels; Rician channels; correlation methods; differential detection; differential phase shift keying; error statistics; land mobile radio; maximum likelihood detection; mobile satellite communication; quadrature phase shift keying; π/4-DQPSK; BER performance; Monte Carlo simulation; Rayleigh channels; correlated Rician fading channel; fading rates; four-symbol detection; ideal channel-information; land mobile satellite communication channels; maximum likelihood detection; multiple-symbol differential detection; no channel-information case; pairwise bit error probabilities; two-symbol detection; union bounds;
Journal_Title :
Communications, IEE Proceedings-
DOI :
10.1049/ip-com:20000335