A unified perspective on ARTM tier I waveforms - part II: common detectors

In the part I companion paper we showed that the tier 1 waveforms FQPSK-JR and SOQPSK-TG - can both be viewed as either an XTCQM or a CPM (although the CPM representation of FQPSK-JR is only an approximation). In this paper we examine the performance of three architectures for a common detector: a symbol-by-symbol detector, an XTCQM detector, and a CPM detector. The symbol-by-symbol detector suffers a 2.0 dB loss relative to the optimum detector for SOQPSK-TG and a 2.2 dB loss for FQPSK-JR when an integrate and dump filter is used and a 1.5 dB loss for SOQPSK-TG and a 1.6 dB loss for FQPSK-JR when a detection filter matched to the average waveform is used. The common XTCQM detector gives a bit error rate performance that is 0.1 dB worse than optimum for SOQPSK-TG and approximately equal to optimum for FQPSK-JR. The common CPM detector gives a bit error rate performance that is 0.25 dB worse than optimum for SOQPSK-TG and about 0.05 dB worse than optimum for FQPSK-JR. The common XTCQM detector provides the best overall performance among the candidate common detectors, but this detector has the highest complexity of the three