Turbo Code Performance in Partial-band Jamming

The Johns Hopkins University/Applied Physics Laboratory Laurel, Maryland ABSTRACT Monte Carlo simulations model the eflects of partial-band jamming on turbo code performance over a slow frequency-hopped spread spectrum signal. Decoders using both hard and soft decision variables corrupted by Additive White Gaussian Noise (A WGN) are considered. Modulations include both Binary Phase Shzjl Keying (BPSK) and dl~erentially coherent BPSK (DPSK). The Bit Error Rate (BER) performance is measured in channels with negligible levels of thermal noise and in channels with similar levels of jammer and thermal noise. The optimum jammer occupancy is measured and found to be consistent with theoretical predictions. Decoder performance is improved by using side-information or by clipping the decision variable. Optimum clipping levels are estimated.