The advancement of technology has resulted in a substantial rise in the number of computing devices and the volume of data being transmitted over networks. The need for fast and secure data encryption has become imperative in response to the increase in data transmission and computing devices. In our previous work, we presented a Fisher-Yates Shuffling (FYS) based image encryption algorithm with a timeout feature that ensures improved security and privacy, regardless of key size. However, the implementation was sequential, and it did not fully utilize the multi-core architecture available on modern computer systems. Therefore, this paper seeks to optimize the FYS-based image encryption algorithm’s performance by parallelizing it on a CPU, with the aim of improving its speed without compromising its security and privacy features. The use of Joblib and multithreading are employed to generate the SHA keys, with a quad-core processor with eight logical processors utilized for the research. The parallelization approach has been tested over thousands of images and has been shown to improve the encryption speed by 2 to 5 times compared to the FYS-based image encryption algorithm. The results demonstrate that using CPU parallelization significantly increases the performance of the FYS-based image encryption algorithm.

Virtualization is widely adopted by the data centers, in order to fulfill the high demand for resources and for their proper utilization. For system management in these virtualized data centers virtual machine live migration acts as a key method. It provides significant benefit of load-balancing without service disruption. Along with the various benefits virtual machine live migration also imposes performance overhead in terms of computation, space and bandwidth used. This paper analyzes the widely used precopy method for virtual machine live migration and proposes the two fold optimization of precopy method for virtual machine live migration. In the first phase, the proposed two fold precopy method reduces the amount of data sent in first iteration of precopy method. Second phase restricts sending of similar data iteratively in each subsequent iterations of precopy method by identifying frequently updated pages and keeps it till the last stop and copy iteration. In this way it reduces total migration time and total amount of data transferred. The proposed two fold precopy method is compared with precopy method and simulation results show the performance improvement of a virtual machine live migration in terms of total migration time and total amount of data transferred.