Optimum threshold signal detection in broadband impulsive noise employing both time and spatial sampling

This paper investigates the optimum detection of a weak signal, where this signal is dominated by broadband impulsive noise. It has been suggested that by using a form of space diversity reception (many receiving sites), it would be possible to 'look between the large noise pulses' and thereby detect a signal well buried in the noise. In this paper we obtain and determine the performance of the locally optimum (or threshold) detector for any form of interference; apply these results to an example case of atmospheric noise for the single receiver case and for the diversity case of many receivers; and then present various numerical results. Substantial improvement or processing gain can be obtained with one receiver, with substantial, but less, additional improvement obtainable by then going to many additional receivers.