An Adaptive Receiver for Signals of Unknown Frequency

The problem of detection of signals of unknown frequency is analytically described and statistical decision theory is applied to determine the receiver that optimally uses past data to improve present performance. The theoretical peformance of this optimum adaptive receiver is compared with two nonadaptive receivers (the energy detector and an optimum nonadaptive receiver) and it is found that in some cases theoretical performance improvement may be worth the extra complexity of the adaptive receiver. Since such receivers have not been constructed, the results of tests on a feasibility model recently completed are also reported. These tests indicate that theoretical performance improvement is so difficult to attain practically, that adaptive receivers of signals of unknown frequency will probably not be economically useful for several years except in those cases where suitably fast generalpurpose computers are available.