EC701

Detection and Estimation    3-0-0-3

COURSE OBJECTIVE

  • The objective of this course is to make the students conversant with those aspects of statistical decision and estimation which is indispensable tools required for the optimal design of digital communication systems.

 

COURSE CONTENT

Binary hypothesis testing; Bayes, minimax and Neyman-Pearson tests. Composite hypothesis testing.

Signal detection in discrete time: Models and detector structures. Coherent detection in independent noise. Detection in Gaussian noise. Detection of signals with random parameters. Detection of stochastic signals. Performance evaluation of signal detection procedures.

Bayesian parameter estimation; MMSE, MMAE and MAP estimates. Nonrandom parameter estimation. Exponential families. Completeness theorem. ML estimation. Information inequality. Asymptotic properties of MLEs.

Discrete  time  Kalman-  Bucy  filter.  Linear  estimation.  Orthogonality  principle.  Wiener-  Kolmogorov filtering – causal and noncausal filters.

Signal detection in continous time:Detection of deterministic signals in Gaussian noise. Coherent detection in white Gaussian noise.

 

Text Books

1.   H.V.Poor, “An Introduction to Signal Detection and Estimation (2/e) Springer”, 1994.

2.   B.C.Levy,  “Priciples of Signal Detection and Parameter Estimatio”n, Springer, 2008.

 

Reference Books

1.   H.L.Vantrees, “Detection, Estimation and Modulation theory”, Part I, Wiley,1987.

2.   M.D.Srinath & P.K.Rajasekaran, “Statistical Signal Processing with Applications”, Wiley, 1979.

3.   J.C.Hancock & P.A. Wintz, “Signal Detection Theory”, Mc-Graw Hill, 1966.

 

COURSE OUTCOMES

Students are able to

CO1: summarize the fundamental concept on Statistical Decision Theory and Hypothesis Testing

CO2: summarize the various signal estimation techniques with additive noise

CO3: summarizer with Bayesian parameter estimation (minimum mean square error (MMSE), minimum mean absolute error (MMAE), maximum a-posterior          probability (MAP) estimation methods).

        CO4: compare optimal filtering, linear estimation, and Wiener/Kalman filtering.

        CO5: construct Wiener and Kalman filters (time discrete) and state space models.