No abstract

This paper analyzes the similarity between steganography and adversarial attack, and then analyzes the attack and defense stage of adversarial samples from the perspective of steganography, which provides a broader idea for researchers in this field. This paper validates the efficiency of steganography on help resisting adversarial attack and proposes a steganalysis based method to enhance the adversarial attack against steganalysis detectors with the help of a distortion function. We combine the distortion map with gradient information to select the disturbed pixels in each attack iteration to generate adversarial samples that can bypass steganalysis detectors. The results show that steganography technology can help detecting normal adversarial samples, and our enhanced adversarial samples achieve a much higher transparency rate against our steganography based method than other normal adversarial samples.

For the traditional distributed computing system, the related message logging conditions had specified to keep the state consistence among the distributed process. Since many new characteristics are introduced in the mobile computing system, the new sufficient logging requirement for mobile computing has to be specified, to avoid the possible state inconsistence among the mobile nodes and the static nodes during the rollback recovery. Firstly, the related definitions of the inconsistence among the process state and the nondeterministic event were extended in the paper, independent of the specific log-based fault tolerant scheme. Finally, a novel particular logging consistency condition was derived based on the extended definitions and Piece-Wise Deterministic model for the mobile computing system. By contrast, the proposal is a practical and efficient constraint for mobile computing upon the possible failures.

No abstract

Many new characteristics are introduced in the mobile computing system, such as mobility, disconnections, finite power source, vulnerable to physical damage, lack of stable storage. Many log-based rollback recovery fault tolerant schemes were proposed according to the characteristics. However, these schemes may still lead to dramatic loss of computing performance in failure-free or inconsistent recovery after the process fault. In this paper, a hybrid log-based fault tolerant scheme is proposed combining the checkpointing mechanism with the message logging mechanism. The checkpoint, the logs and the happened-before relations are logged synchronously into the memory at local mobile hosts temporarily, and asynchronously into the persistent disk storage in the form of the antecedence graph at local mobile support station. The proposal supports the independent and propagated consistent recovery. By contrast, the results show that the proposal incurs a lower failure-free overhead on the premise of the consistent recoverability.