Optimum quantization for detection

Picinbono, B.; Duvaut, Patrick

doi:10.1109/26.8934

Cited by 47 publications

(21 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Section 4, we establish the optimality of LRQs when the performance criterion is an AliSilvey distance mesure [AS66, PT771, thus providing a broad generalization of the results of [PD88, KVW891.…”

Section: Summary Of the Papermentioning

confidence: 99%

“…Independently from [FG87], the optimality of LRQs was established in [PD88] for the special case f(z) = l/z, using a direct argument. Following the lines of the argument in [PD88], [KVW89] the fusion center solves a Neyman-Pearson hypothesis testing problem to decide in favor of one of the underlying hypotheses. When N is large, the probability of error by the fusion center can be approximated by e -N F(7), where F is defined by Eq.…”

Section: Proposition 41: the Problem Of Finding A Quantizer That Maxmentioning

confidence: 99%

See 1 more Smart Citation

Extremal properties of likelihood-ratio quantizers

Tsitsiklis

1993

IEEE Trans. Commun.

107

118

View full text Add to dashboard Cite

Let there be M hypotheses H 1 ,... , HM, and let Y be a random variable, taking values in a set y, with different probability distribution under each hypothesis. A quantizer -': Y-+ {1,..., D} is applied to form a quantized random variable 7(Y). We characterize the extreme points of the set of possible probability distributions of -(Y), as 7y ranges over all quantizers. We then establish optimality properties of likelihood-ratio quantizers for a very broad class of quantization problems, including problems involving the maximization of an Ali-Silvey distance measure. Some new results are also obtained for a Neyman-Pearson decentralized detection problem.

show abstract

Section: Summary Of the Papermentioning

confidence: 99%

Section: Proposition 41: the Problem Of Finding A Quantizer That Maxmentioning

confidence: 99%

Extremal properties of likelihood-ratio quantizers

Tsitsiklis

1993

IEEE Trans. Commun.

107

118

View full text Add to dashboard Cite

show abstract

“…Note that the cardinality of An is same as the cardinality of Ln. Therefore the set A, which is the set of values Tsd can take, contains all possible combinations of sums of values in A n for n = 1, ... , N: (14) The distribution of TN can be evaluated using the convolution of the pmfs of the discrete independent random variables An …”

Section: A Sequential Detectionmentioning

confidence: 99%

Effects of quantization and channel errors on sequential detection in cognitive radios

Chaudhari

Lundén

Koivunen

2012

2012 46th Annual Conference on Information Sciences and Systems (CISS)

View full text Add to dashboard Cite

This paper studies the impact of quantization and reporting channel errors on the performance of a sequential detection (SD) scheme for cooperative spectrum sensing in cognitive radios. The analyzed system model is a distributed parallel detection network in which the sensors quantize their local log-likelihood ratios (LLRs) and transmit them to the FC which makes the decision sequentially. The quantized local LLRs are transmitted through erroneous reporting channels. These reporting channels are modeled as binary symmetric channels (BSCs) that cause errors with a certain bit error probability (BEP). The quantization and reporting channel errors are shown to increase the average sample number (ASN) of the SD. Later the performances of the SD and fixed sample size (FSS) test are compared to show the superiority of the SD over the FSS test even in the presence of quantization and reporting channel errors regardless of the BEP. Even the ASN of the SD with I-bit quantization is shown to be smaller than the ASN of an unquantized FSS test with the same error probabilities in the considered scenario.

show abstract

“…Various properties of such systems are analyzed in [7] and extensions to higher order properties are given in [8 and 101.…”

Section: A Advantages Of the Deflectionmentioning

confidence: 99%

On deflection as a performance criterion in detection

Picinbono

1995

IEEE Trans. Aerosp. Electron. Syst.

Self Cite

120

View full text Add to dashboard Cite

The performance of detection system is usually characterized by the receiver operating characteristics (ROC) curves. Because of the complexity of the calculation of such curves simpler performance criteria are useful. Among these, the deflection criterion, or output signal-to-noise ratio, is one of the most interesting. The advantages and limitations of the deflection are first analyzed. Attention is especially paid to the relation between singular detection and S i n i t e deflection Maximization of the deflection in the class of linear-quadratic (L-Q) system is also discussed and it is shown that in order to detect a deterministic signal in some kinds of non-Gaussian noise a quadratic term can introduce an almost singular detection This situation is analyzed by computer simulations and comparisons between results obtained via the deflection or the ROC curves indicate the interest of the deflection in the analysis of the performance of some nonlinear system.

show abstract

Optimum quantization for detection

Cited by 47 publications

References 10 publications

Extremal properties of likelihood-ratio quantizers

Extremal properties of likelihood-ratio quantizers

Effects of quantization and channel errors on sequential detection in cognitive radios

On deflection as a performance criterion in detection

Contact Info

Product

Resources

About