Hudan Studiawan

Shared secret-based steganography for protecting medical data

Medical data (i.e., medical image and medical record) is private which should be protected before being transferred or stored. There have been some methods proposed by implementing various algorithms, such as steganography. It is sometimes combined with the shared secret method, so that, the reconstruction of the original data must be done by some users. It is believed that this method increases the security of the data. However, there is still a need to make the stego data (i.e, stego image) has an as good as possible quality. This means that the stego image has high similarity to the cover (carrier) image. In this paper, we propose methods to protect the medical data by using the shared secret mechanism and steganography for 2 and 1 bit LSB. The experimental results show that they produce relatively good quality (similarity) in terms of PSNR.

Image thresholding using ultrafuzziness optimization based on type II fuzzy sets

Image thresholding is a critical process in digital image processing application. However, there are some disturbing factors like image vagueness and bad illumination resulting in not satisfied image thresholding output. Several fuzzy thresholding techniques are developed to remove graylevel ambiguity during threshold selection. One of the techniques is thresholding method using type II fuzzy sets. In this paper, we propose relaxation of the ultrafuzziness measurement by considering ultrafuzziness for background and object fuzzy sets separately. The proposed method optimizing ultrafuzziness to decrease uncertainty in fuzzy system used type II fuzzy sets. Experimental results on several images show the effectiveness of the proposed method.

SQL injection detection and prevention system with raspberry Pi honeypot cluster for trapping attacker

One of the most common security attack for web application is SQL injection. It is an attack to acquire access to applications database through injection of script or malicious query attributes. This attack can be executed in any page of web application which interacts with database. SQL injection could be more dangerous if the victim was an enterprise system such as online banking. Many methods have been researched and developed to prevent SQL injection attacks. One of them is the use of a honeypot. This paper proposed a method for increasing systems capability to detect and prevent SQL injection attacks based on removal of SQL query attribute values and honeypot for trapping attackers. A honeypot is placed as decoy system to hide actual web server from attacker. Malicious queries from attackers will be sent to honeypot while normal queries will be sent directly to the real web server. Honeypot is also used to provide activity logging of each attack which can be used for further analysis. We play with Raspberry Pi because it is cheap and effective to be used as a honeypot. Due to its limited computational ability, we make cluster to improve its power. Based on conducted experiments, we could achieve up to 64% accuracy of SQL injection attack. Moreover, with the redirection, our honeypot could get more attack data to be analyzed.

Adaptive image compression using Adaptive Huffman and LZW

In this digital era, the need of storing data has increased rapidly. This circumstance is proportional to the size of files and their storage. In order to decrease the required big size of storage, the file size should be reduced by still considering the quality of the respective data. This can be done by implementing a compression algorithm, such as LZW. In this paper, we propose the pre-processing step which is used before the file is being compressed. This step includes bit selection by using mean, median, and mode for adaptively determining the number of replacing bits. According the experimental result performed to the standard test images, we are able to achieve 36.26 dB of PSNR and 2.9 of compression ratio.

Rancang Bangun Sistem Navigasi Indoor Berbasis Integrasi Symbolic Location Model Dan Wifi Based Positioning System Untuk Studi Kasus Pada Gedung Bertingkat

Kemajuan teknologi informasi saat ini memudahkan setiap orang dalam kegiatannya masing-masing. Saat ini melakukan navigasi tidak perlu menggunakan selembar peta lagi tetapi cukup mengakses GPS ataupun penyedia jasa navigasi lainnya. Teknologi navigasi saat ini sudah sangat maju untuk navigasi di luar ruangan tetapi untuk di dalam ruangan, belum ada standar khusus dan belum terlalu umum digunakan. Tempat umum di Indonesia, khususnya gedung bertingkat, belum menggunakan teknologi sebagai alat navigasinya dan hanya menggunakan selebaran atau peta denah saja. Saat ini masyarakat sudah banyak yang memiliki smartphone sehingga mudah untuk berbagi atau mencari tempat yang diinginkan. Dalam penelitian ini, Indoor Positioning berbasis Wi-Fi akan digunakan untuk menentukan posisi menggunakan kekuatan sinyal beberapa Wi-Fi yang berbeda dengan algoritma K Nearest Neighbor dan Euclidean Distance. Symbolic Model digunakan untuk memodelkan sebuah gedung bertingkat dan dapat menangani kasus-kasus yang ada dalam memodelkan lokasi, seperti overlap, mengandung dan terhubung. Dari hasil pengujian, pengguna dapat melakukan pencarian posisi dengan tingkat akurasi sebesar 86.66 dan dapat melakukan navigasi ke setiap lokasi yang ada di Jurusan Teknik Informatika

Clustering of SSH brute-force attack logs using k-clique percolation

The brute-force attacks to SSH service still persist in the server environments. The existing methods have not applied graph theory to analyze authentication log that records this attack. Therefore, we model the log as a graph and propose k-clique percolation to cluster auth.log file to assist the system administrators to inspect this incident. The k-clique percolation has proven in clustering of biological networks and we will deploy it to this problem. We then provide the mechanism for edge removal to separate the generated clusters and make clear the clustering outputs. The experimental results show that this approach is appropriate to cluster raw logs of SSH brute-force attacks.

Emoticon-based steganography for securing sensitive data

Transferring information between nodes in this digital era has been made it easy for users to communicate. Sensitive data, however, requires protection during transmission to prevent unauthorized users from exploiting it. In this paper, we propose to hide secret data by developing emoticon-based steganography. Here, the data is embedded to the emoticon which is sent along with the cover text via chatting. The experimental result, which is obtained from various types of data, depicts that this proposed method, in general, has higher performance than that of existing one. That is, each emoticon has been able to dynamically hold more information. Specifically, the method without preprocessing is better.