Peter Truchly

Syn Flood Attack Detection and Type Distinguishing Mechanism Based on Counting Bloom Filter

Presented work focuses onto proposal, implementation and evaluation of the new method for detection and type identification of SYN flood (DoS) attacks. The method allows distinguishing type of detected SYN flood attacks – random, subnet or fixed. Based on Counting Bloom filter, the attack detection and identification algorithm is proposed, implemented and evaluated in KaTaLyzer network traffic monitoring tool. Proof of correctness of the approach for TCP SYN flood attack detection and type identification is provided – both in practical and theoretical manners. In practice, new module for KaTaLyzer is implemented and TCP attacks are detected, identified and network administrator is notified about them in real-time.

Enhanced hybrid TV platform with multiscreen, advanced EPG and recommendation enablers

Abstract TV watching dramatically changes with introduction of new technologies such as Internet-connected TVs, enriched digital broadcasting (DVB), on-demand content, additional programme information, mobile phones and tablets enabling multiscreen functions etc that offer added values to content consumers. In this paper we propose modular advanced TV platform and its enablers enhancing TV watching. They allow users to receive aside of EPG also additional information about broadcasted content, to be reminded of requested programme, to utilize recommendation and search features, thanks to multiscreen functionality to allow users to take watched content with them or transfer it onto another device. The modularity of the platform allows new features to be added in future.

Security aspects in education oriented multimedia networks

The article summarizes security aspects related to multimedia networks with special focus to user identification, authentication and authorization procedures and mechanisms. Main stress is taken into account to the secure and flexible access of end users to multimedia contents, implemented on the application layer of the ICT architecture platform. The novel network architectures supported innovative educational paradigms “Student self-directed learning” based on “Student centric” model of education, is described in the paper. Practical part of the paper introduces application of modern multimedia network architecture, with the integration of the actual HbbTV concept in the heterogeneous ICT infrastructure, to provide online and off-line educational multimedia content to end users - students participating in the e/m-learning process. The Service Oriented Architecture (SOA) based on web services, a “distributed architecture that enables software components to be published as services over the Internet through a programmable interface and using Internet protocols for the purpose of the providing ways to find, subscribe, and invoke those services” [4]. The wide portfolio of the multimedia content and e/m-learning approaches and mechanisms are presented in the paper. The integration of the HbbTV concept to the heterogeneous ICT infrastructure, applicable for the on-line and off-line educational process is described in the paper.

Acoustic and auxiliary speech features for speaker identification system

The focus of the article is on the selection, adjustment and overall performance of speech features at acoustical and prosodic level for speaker recognition task. Namely: perceptual linear prediction, Mel frequency cepstra, cepstral linear prediction, formant frequencies, and different auxiliary features. Both brief theoretical backgrounds and possible computational methods are outlined in regard to the speaker recognition task. In the series of experiments using 114 speakers database, it was observed that a model based method slightly outperformed the perceptual ones. Furthermore, it was found that auxiliary and prosodic features may not always improve scores when processed together with acoustic ones. On average the success rate was about 90% whereas the best recorded score was 99.1% for cepstral linear prediction coefficients in connection with k-nearest neighbor classifier.

Multilingual and multisensorial dictionary tool as a support for the effective learning in the area of ICT

E-learning can be more efficient when training materials are supported by various additional interactive and explanatory tools. One such supporting tool we have developed and implemented in the framework of several Leonardo da Vinci (LdV) and Erasmus+ projects focused on the lifelong learning of experts from ICT area and vocational training. Training materials for topics in information and communication technologies are full of technical terminology and abbreviations. Therefore we provided students with a multilingual and multisensorial terminological dictionary. This paper presents the evolution, modifications and enhancements of this tool throughout the four EC projects. We mainly concentrate on its basic functions and features as well as graphical user interfaces which offer the end users the flexible access to multilingual terminological database. We also explain reasons why we changed its nature in particular projects.

IP Data Delivery in HBB-Next Network Architecture

Digital television enables IP data delivery using various protocols. Hybrid television HbbTV enhances digital television with applications delivery. HBB-Next is an architecture which enhances HbbTV with additional features. However it does not specify IP data delivery despite it has access to both broadcast and broadband channel. This paper proposes architecture and protocols for IP data delivery in HBB-Next. To achieve this goal we designed new node (Application Data Handler - ADH) in HBB-Next architecture and new communication protocols (Application Data Handler Control Protocol - ADHCP, and Hybrid Encapsulation Protocol - HEP) for data transmission. We created Stochastic Petri Net (SPN) model of designed protocols and implemented them in ns2 network simulator to verify our solution. Results of SPN model simulation and ns2 network simulation are discussed and HEP protocol is compared to existing encapsulation protocols used in DVB systems.

Wlan Performance Improvement by Header Compression and Packet Overhearing Reduction

Abstract For wireless systems, it is of vital importance to use the scarce radio resources in an efficient way. The main targets are to reduce the power consumption and increase the throughput of wireless devices. Using header compression both throughput can be increased and power can be saved. Using packet overhearing reduction power can be saved. This paper presents the achieved performance improvements when using these two methods. The simulation results are confirmed by analytical calculations. We show that we can achieve up to 35% power save when combining both methods and close to 0.1 additional normalized throughput.