Stefan Pfennig

Spectral methods to determine the exact scaling factor of resampled digital images

This paper combines analytical models of periodic interpolation artifacts with recent empirical findings on the spectral energy distribution of rescaled images to infer exact transformation parameters in a passive-blind forensic setting. We present a measure to solve a long-known ambiguity between upscaling and downscaling in the forensic analysis of resampled signals and thus substantially limit the range of candidate scaling factors. The effectiveness of our method is backed with empirical evidence on a large set of images and scaling factors.

Physical layer security vs. network layer secrecy: Who wins on the untrusted two-way relay channel?

We consider the problem of secure communications in a Gaussian two-way relay network where two nodes exchange confidential messages only via an untrusted relay. The relay is assumed to be honest but curious, i.e., an eavesdropper that conforms to the system rules and applies the intended relaying scheme. We analyze the achievable secrecy rates by applying network coding on the physical layer or the network layer and compare the results in terms of complexity, overhead, and efficiency. Further, we discuss the advantages and disadvantages of the respective approaches.

Efficiency of secure network coding schemes

Network coding is a promising approach for increasing performance of multicast data transmission and reducing energy costs. Of course, it is essential to consider security aspects to ensure a reliable data transmission. Particularly, pollution attacks may have serious impacts in network coding since a single attacker can jam large parts of the network. Therefore, various approaches have been introduced to secure network coding against this type of attack. However, introducing security increases costs. Even though there are some performance analysis of secure schemes, to our knowledge there are no details whether these schemes are worthwhile to replace routing under the facet of efficiency. Thus, we discuss in this paper parameters to assess the efficiency of secure network coding schemes. Using three network graphs, we evaluate parameters focusing on communication overhead for selected schemes. Our results show that there are still benefits in comparison to routing depending on the network topology.

Modeling communication delays for network coding and routing for error-prone transmission

During the execution of parallel applications, there might be a very large number of exchanged messages between the computing nodes. Hence, the communication delays crucially influence the execution time of the whole application and a suitable communication approach should be selected to minimize the execution time. Selecting a suitable communication scheme requires a model that allows to compare communication delays of potential communication schemes considering system param-eters, such as network topology or specific processing delays. Within this paper, we present an approach for modeling the communication delays depending on system parameters such as delay for sending and receiving data. The considered communication schemes are network coding and routing. We present a model for the influence of the multiple delays caused by transferring the data and formulas for the case of error free transmission. For the error-prone case, we conduct discrete event simulations that deliver communication delays depending on system parameters as well as on error probabilities. Our model for communication delays can be used to select a suited communication scheme in order to reduce the overall execution time of a parallel application.

Adjustable redundancy for secure network coding in a unicast scenario

Within this paper, we investigate efficient communication by means of network coding in a unicast scenario. Packets can be lost due to errors or attacks, hence, redundant data packets are necessary to ensure a successful transmission. In the ideal case, the level of redundancy should correspond to the given loss rate. However, the loss rate may be unknown in practice. We investigate methods for determining the number of redundant packets based on estimating the delivery rate. The estimation utilizes feedback by the recipient. Each of the investigated methods is characterized by a certain trade-off between communication overhead and delay. Depending on system conditions and application requirements, these goals can have different importance. Thus, we introduce a method for adjusting the redundancy according to individual needs regarding communication overhead and delay. Simulation results confirm that the estimation of the delivery rate is suitable for determining the level of redundancy and that it is possible to adapt to individual optimization goals: minimizing communication overhead and/or delay.

End-to-End Key Establishment with Physical Layer Key Generation and Specific Attacker Models

Physical layer key generation got much attention during the last time. However, the need of a common physical channel implies that only point-to-point keys can be generated. In this chapter, we investigate approaches how these point-to-point keys can be used for a secure establishment of end-to-end keys between two users who can only communicate over a multi-hop network. We start with a review of physical layer key generation taking different attacker models into account. Subsequently, we introduce general approaches for the end-to-end key establishment in the presence of various attackers who differ in their behavior and their area of control. We discuss four different path selection algorithms for the key establishment and evaluate their performance by means of simulations. The results show that the end-to-end key establishment can be protected by means of physical layer keys with a reasonable effort if suitable path selection is applied.

Confidential Network Coding: Physical Layer vs. Network Layer

In all kind of information exchange, security is essential. One protection goal that has to be enforced is confidentiality. In state-of-the-art protocols, messages are encrypted before they are transmitted to ensure their confidentiality. However, incorporating novel technologies like network coding allows for more efficient solutions. Within this article, we compare different solutions for confidential communication by means of network coding at physical layer and at network layer. We discuss security, efficiency, and computational complexity of these approaches. The results allow to draw conclusions about the choice of a suited communication scheme depending on the system model and the relevant parameters.

Energy Models for Communication of Future Computing Platforms

Energy efficiency is an essential requirement on future computer architectures. Thereby, the efficiency of the communication will play an important role. Within this paper, we present an energy model for a unicast multi-hop communication. The model allows to derive predictions about the energy of a future computing platform. Further, it can be used to identify which components of the system or tasks related to communication have a significant influence on the overall energy and, hence, require special attention in the design phase. We discuss how values for processing delays and power consumption can be determined for a future computing platform and present values for a target platform. The energy model is applied to derive energy predictions for this platform.

Analysis of Parallel Applications on a High Performance-Low Energy Computer

In this paper, we propose a holistic approach for the analysis of parallel applications on a high performance–low energy computer (called the HAEC platform). The HAEC platform is currently under design and refers to an architecture in which multiple 3-D stacked massively parallel processor chips are optically interconnected on a single board and multiple parallel boards are interconnected using short-range high-speed wireless links. Although not exclusively targeting high performance computing (HPC), the HAEC platform aims to deliver high performance at low energy costs, which are essential features for future HPC platforms. At the core of the proposed approach is a trace-driven simulator called haec_sim which we developed to simulate the behavior of parallel applications running on this hardware. We investigate several mapping layouts to assign the parallel applications to the HAEC platform. We concentrate on analyzing the communication performance of the HAEC platform running parallel applications. The simulator can employ two communication models: dimension order routing (DOR) and practical network coding (PNC). As a first example of the usefulness of the proposed holistic analysis approach, we present simulation results using these communication models on a communication-intensive parallel benchmark. These results highlight the potential of the mapping strategies and communication models for analyzing the performance of various types of parallel applications on the HAEC platform. This work constitutes the first step towards more complex simulations and analyses of performance and energy scenarios than those presented herein.

Secure network coding: Dependency of efficiency on network topology

Network Coding is a new possibility to transmit data through a network. By combining different packets instead of simply forwarding, network coding offers the opportunity to reach the Min-Cut/Max-Flow capacity in multicast data transmissions. However, the basic schemes are vulnerable to so-called pollution attacks, where an attacker can jam large parts of the transmission by infiltrating only one bogus message. In the literature we found several approaches which aim at handling this kind of attack with different amounts of overhead. Though, the cost for a specific secure network coding scheme highly depends on the underlying network. The goal of this paper is on the one hand to describe which network parameters influence the efficiency of a certain scheme and on the other hand to provide concrete suggestions for selecting the most efficient secure network coding scheme considering a given network. We will illustrate that there does not exist “the best” secure network scheme concerning efficiency, but all selected schemes are more or less suited under certain network topologies.