Military communication systems do not necessarily operate within FCC frequency bands. Hence, they may be subject to interference from other waveforms using the same frequency band. In this paper we first investigate a technique to estimate the spectrum of competing signals utilizing the same bandwidth as a desiredfrequency-hopped waveform. Next, we show that the desired frequency-hopped waveform can be recoveredfrom the composite received signal by dividing the composite signal spectrum by an estimate of the interference spectrum. Since the interference estimate is imperfect, spectral division is significantly better than spectral subtraction of the interference spectrum from the composite spectrum for the detection offrequency-hopped waveforms.

No abstract

No abstract

In this paper the development of the digital ad-hoc chip rate estimator (ACRE) is explained in relation to its similarities and contrasts to the chip rate detector (CRD). The effect of half-sine pulse shaping on chip rate detection is investigated. The components of the ACRE are described. The ACRE's performance for three categories of simulations are presented and briefly discussed. The three simulation categories are: standard-ACRE, ACRE with additional filtering, and ACRE with incrementing. ACRE with additional filtering trades chip rate search range for estimation performance. ACRE with incrementing combines subsets of processed data that overlap each other. This simulation approach shows the potential for parallel processing without loss of performance.

We investigate the ability offinite impulse response (FIR), and infinite impulse response (IIR) filters to remove narrowband interference from afrequency-hopped signal. It is possible to construct narrowband IIR filters with very few taps, however, these filters have a non-linear phase response, and Monte-Carlo simulations demonstrate that this increases the probability of bit error. FIR filters have a linear phase response, but require several orders of magnitude more taps to produce the same narrowband filter frequency response. Finally we consider forward and reverse processing with the IIR filter. By filtering the signal with an IIR filter and then time-reversing the signal andfiltering it again, it is possible to have linear phase at the expense of an increase in processing complexity. Monte-Carlo simulations demonstrate that forward and reverse processing significantly reduces the probability of bit error when IIRfilters are used.