ERROR FLOOR OF UNEQUALIZED WIRELESS PERSONAL COMMUNICATIONS-SYSTEMS WITH MSK MODULATION AND TRAINING-SEQUENCE-BASED ADAPTIVE SAMPLING

We analyze the error door of unequalized personal communications syste ms using minimum-shift keying (MSK) with training-sequence-based adapt ive sampling in a two-delay fading channel, We include Gaussian filter ing of the input, finite accuracy in determining the optimum sampling time, and filtering at the receiver. We prove that for pure MSK and lo w time dispersion, training-sequence-based adaptive sampling can compl etely avoid errors caused by intersymbol interference, The actual erro rs are caused by ''secondary'' effects (filtering and finite-resolutio n sampling) in conjunction with the channel time dispersion, Errors oc cur if the normalized phasor of the channel impulse response falls int o certain error regions; the computation of these regions gives physic al insights into the error mechanism and allows a highly efficient com putation of the average bit-error rate (BER), The average BER varies a s K . (S/T)(2), where S is the rms delay spread and T is the bit lengt h, The proportionality constant K depends on the Gaussian filtering in the transmitter, the receiver filtering, and the amount of oversampli ng. The BER can be orders of magnitude lower than for the (quasi-) fix ed sampling case, in which timing is derived solely based on the chann el delay power profile, For two-branch diversity reception, we shaw th at the BER is proportional to (S/T)(4).