In this work we analyze achievable random coding error exponents pertaining to erasure/list decoding for channels with side information present at the transmitter. The analysis is carried out using the optimal decoding rule proposed by Forney. A key ingredient in the analysis is the evaluation of moments of a certain distance enumerator. This approach leads to a new exponentially tight bounds. These results are obtained by exploring a random binning code with conditionally constant composition codewords previously proposed by Moulin and Wang. Later, these results are used to obtain an achievable random coding error exponent for ordinary decoding.

We consider a decoder with an erasure option and a variable size list decoder for channels with non-casual side information at the transmitter. First, a universally achievable region of error exponents is offered for decoding with an erasure option using a parameterized decoder in the spirit of Csiszár and Körner's decoder. Then, the proposed decoding rule is generalized by extending the range of its parameters to allow variable size list decoding. This extension gives a unified treatment for erasure/list decoding. An achievable region of exponential bounds on the probability of list error and the average number of incorrect messages on the list are given. Relations to Forney's and Csiszár and Körner's decoders for discrete memoryless channel are discussed. These results are obtained by exploring a random binning code with conditionally constant composition codewords proposed by Moulin and Wang, but with a different decoding rule and a modified analysis.

We consider a decoder with an erasure option and a variable size list decoder for channels with non-causal side information at the transmitter. First, universally achievable error exponents are offered for decoding with an erasure option using a parameterized decoder in the spirit of Csiszar and Korner's decoder. Then, the proposed decoding rule is generalized by extending the range of its parameters to allow variable size list decoding. This extension gives a unified treatment for erasure/list decoding. Exponential bounds on the probability of list error and the average number of incorrect messages on the list are given. Relations to Forney's and Csiszar and Korner's decoders for DMC are discussed. These results are obtained by exploring a random binning code with conditionally constant composition codewords proposed by Moulin and Wang, but with a different decoder.

An information-theoretic approach is proposed to watermark embedding and detection under limited detector resources. First, the attack-free scenario is considered under which asymptotically optimal decision regions in the Neyman-Pearson sense are proposed, along with the optimal embedding rule. Later, the case of zero-mean independent and identically distributed (i.i.d.) Gaussian covertext distribution is explored with unknown variance under the attack-free scenario. For this case, a lower bound on the exponential decay rate of the false-negative probability is proposed. It is proven that the optimal embedding and detecting strategy is superior to the customary linear, additive embedding strategy in the exponential sense. Finally, these results are extended to the case of memoryless attacks and general worst case attacks. Optimal decision regions and embedding rules are offered, and the worst attack channel is identified.

We revisit the problem of optimal watermark embedding and detection for computationally limited detectors under general attack channels. First, a formal definition of the problem is given for general attack channels under the Neyman-Pearson criterion of optimality. Later, asymptotically optimal decision regions are offered, assuming that the detector is computationally limited and that the attack channel is strongly exchangeable. These results are extended to general attack channels where random watermark sequences are considered. The optimal embedding rule is offered and the worst attack channel is identified.

An information-theoretic approach is proposed to the watermark embedding and detection under limited detector resources. First, asymptotically optimal decision regions are presented in the Neyman-Pearson sense. These results are also modified to the case of zero-mean i.i.d. Gaussian covertext distribution with unknown variance. For this case, a lower bound is offered on the exponential decay rate of the false-negative probability and it is proven that the optimal embedding and detecting strategy is superior to the customary linear, additive embedding strategy in the exponential sense

In this paper, an efficient procedure for designing a sequence of predictors with memory size equal to the Markov order of source, with error exponent arbitrarily close to the optimal error exponent is presented. In this procedure the moment-generating function of the prediction loss is recursively minimized.