Abstract:The paper presents a new efficient H.264/AVC 4 × 4 intraprediction scheme. The new prediction scheme is based on the best prediction matrix mode. The main idea behind the new prediction scheme is to combine the most usable intraprediction modes, {vertical -horizontal -DC} , into a new efficient prediction mode. The new prediction scheme is implemented using VHDL and hence it uses the full advantages of inherent parallelism in the hardware. We evaluate the performance of this prediction scheme in terms of compression ratio, peak signal to noise ratio, and bit rate using seven video sequences. Moreover, we analyze the power consumption, the delay, and FPGA area utilization of the implemented H.264 encoder after utilizing the new prediction scheme. The performance measures as well as the area and power consumption are compared to other best known prediction algorithms.
Exponential increase in data inside endeavors is making a gush in the storage requirements and its security. Typical encrypting techniques of separable nature suffer against powerful adversaries due to the exponential improvement of computing power. Therefore, some cloud storage systems apply erasure coding in addition to encryption to prevent adversaries from revealing or/and controlling stored information. Nevertheless, still a powerful adversary can extract useful information from a compromised server. In this paper, a new storage scheme is introduced implying an AllOr-Nothing (AON) encryption mode to separately randomize the encrypted blocks and hashing the resulted pseudorandom blocks altogether afterwards, and with the aid of the salting technique, called Replicated BYTE, integrity is ensured. The scheme will be called Randomized AON plus Replicated BYTE (RAON-RB). By employing systematic Reed Solomon erasure coding, the proposed scheme will be more applicable to distributed systems. The proposed scheme is secure even if all but one storage servers have been compromised and even if the encryption key is disclosed. Moreover, to resist share modification and localize faulty server(s), the proposed scheme verifies the received shares before they are involved in the reconstruction process, and hence, the scheme can save superfluous computations as well.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.