A simultaneous Mary channel hypothesis test with leastmeansquare signal amplitude estimation
Abstract
This paper presents a new strategy for detecting and estimating multiple linearly independent signals immersed in additive random noise. This new approach is a Maximum Loglikelihood Ratio Test (MLRT) and determines the presence of one or more signals through a single filter operating on a received input data vector. In addition, it performs a signalamplitude estimation when a positive response occurs. This test is a UNiformly Most Powerful Invariant Test and reduces the computationally intensive work to only those situations where important data exists. The estimation technique determines the signal amplitude of all possible input signals in a leastmeansquare sense with a variance inversely proportional to the inherent signaltonoise ratio of each linearly independent library waveform. Computer simulations verify the theory and demonstrate nearoptimum performance of the MLRT for small signal libraries. The MLRT is close to optimum for larger signal sets, and is clearly optimum for complex multiwaveform signals. ratio test(MLRT), and Signaltonoise ratio.
 Publication:

Interim Technical Report
 Pub Date:
 August 1987
 Bibcode:
 1987nosc.reptS....R
 Keywords:

 Computerized Simulation;
 Least Squares Method;
 Maximum Likelihood Estimates;
 Mean Square Values;
 Signal Processing;
 Signal To Noise Ratios;
 Amplitudes;
 Invariance;
 Optimization;
 Strategy;
 Communications and Radar