Performance of fast frequency hopped noncoherent MFSK with a fixed hop rate under worst case jamming

The authors evaluate the throughput performance of FFH/MFSK (fast frequency hopping/M-ary frequency-shift keying) with a fixed hop rate for both uncoded and coded systems. The performance criterion used is a throughput measure, that is information rate sustained by a system for a given bit-error rate normalized by the hop rate. It is shown that the performance can be greatly improved through the use of coding. As with a fixed data rate, multitone jamming is more effective for M>2, and partial-band-noise jamming for M=2. However, the optimum M is a monotonic increasing function of E/sub h//J/sub O/. This affects the choice of a code if it is related to M, such as a dual-K code. The best convolutional codes are the rate 1/2 Trumpis code, the semiorthogonal code for K=3, and the dual-K codes. The best block codes are the (255,221) RS code, the (255,207) BCH code, and the (63,53) RS code, in that order. As with a fixed data rate, the best convolutional codes perform better than the best block codes.<
>