Alvin Huseinovic

Combining static and live digital forensic analysis in virtual environment

Traditional digital forensics is performed through static analysis of data preserved on permanent storage media. Not all data needed to understand the state of examined system exists in nonvolatile memory. Live analysis uses running system to obtain volatile data for deeper understanding of events going on. Sampling running system might irreversibly change its state making collected evidence invalid. This paper proposes combination of static and live analysis. Virtualization is used to bring static data to life. Volatile memory dump is used to enable offline analysis of live data. Using data from memory dump, virtual machine created from static data can be adjusted to provide better picture of the live system at the time when the dump was made. Investigator can have interactive session with virtual machine without violating evidence integrity. Tests with sample system confirm viability of proposed approach.

Forensic analysis of encrypted volumes using hibernation file

Nowadays, software tools are commonly used to encrypt data on hard disk. Those tools keep encryption keys in system memory to provide the user easy access to plain text of encrypted files. Key possesion enables data decryption. A procedure that includes usage of hibernation file as a source of memory content is described. Publicly available tools are used to perform the procedure. The procedure is successfully tested on a system that uses current encryption program.

Virtual machine memory forensics

Physical memory can contain various data such as user passwords, encryption keys, web browser activity and other traces interesting for forensic analysis. Virtual machine physical memory is usually presented as a file on a host operating system. In this paper, the obtaining and analyzing of the virtual machine memory dump are presented.

Utilizing CUDA architecture for improving application performance

Graphic card processors have a vital role in that they bring parallelism to a whole new level. Applications need to do a large amount of calculations in parallel. In this paper numeric calculation performance comparison between executing numerical calculations in traditional processor and the graphics processor using CUDA architecture is presented.

Parallelization challenges of BFS traversal on dense graphs using the CUDA platform

This paper presents challenges encountered while parallelizing an existing sequential algorithm. A breadth-first search implementation in CUDA C++ of quadratic time complexity is used. Even though BFS might seem like an easily parallelizable problem due to many independent iterations over graph vertices, there are other important aspects which need to be considered. Properties like granulation, communication and load balancing are thoroughly examined to find their exact impact in BFS. The implementation is profiled to find bottlenecks and problems which prevent its further optimization.

Benchmark comparison of computing the Mandelbrot set in OpenCL

Computing the fractals can be time consuming if processed in sequential manner. Computing the fractals is iterative process so it can easily be transformed to run in parallel fashion. Since OpenCL supports parallel execution of code on both CPU and/or GPU, benchmarking and comparison the computation of Madelbrot by using OpenCL, can give insight on the speedup that modern parallel architectures have to offer. Results of these benchmarks and comparsions are given in this paper.

Disabling same origin policy for automatization of webform data entry

There is often a need enter data in online documents. Web based interface is popular, but sometimes there is a need to fill the data using scripting and automation. JavaScript may enable automatic data entry, but it requires that script written in JavaScript executes on the same server where the form is being processed, due to protection mechanism of “same origin policy”. If needed, some browsers support disabling this policy, or there is alternative method to bypass it. The paper describes this method and includes a sample JavaScript code for automated data confirmation.

Attempt of unbiased comparison of GPU and CPU performance in common scientific computing

Graphics processing units (GPU) are considered to have superior performance over the central processing units (CPU) in performing common scientific computations. Number of factors that can seriously change this conception are usually overlooked. In this paper, some of these factors are taken into account and their impact is measured, analysed and discussed. Matrix multiplication and Shell sorting examples are used to show that GPU superiority is not always clear.

Implementation and performance analysis of the Simplex algorithm adapted to run on commodity OpenCL enabled graphics processors

The Simplex algorithm is commonly used for solving Linear Optimization problems. Linear Optimization methods are used to solve problems in areas such as Economics, Business, Planning and Engineering. Developments of hardware platforms have allowed the use of Linear Optimization methods on problems that presented serious computational challenges in the past. However, solving large optimization problems can be time consuming, which has to be taken into consideration for time-critical applications. With the invention of the GPU assisted computing the situation in this field has progressed. In this paper, implementation and performance analysis of the Simplex algorithm adapted to take the advantage of modern graphics processors versus traditional CPU adapted implementation is presented.