1,721,140 research outputs found
On the performance of constrained amplify-and-forward networks
Get PDFThis thesis examines the effects of resource constraints on amplify-and-forward
(AF) networks.
Chapters 3 and 4 are the first research chapters. Chapter 3 studies the outage
probability performance of a two-hop two-way AF peak power constrained orthogonal
frequency division multiplexing (OFDM) network. Its performance is then
optimized. Chapter 4 focuses on the one-way special case of the system studied
in Chapter 3. It begins with an analysis of the network when nonlinear distortion
produced by signal clipping dominates the additive noise in the system. To conclude
Chapter 4, the theoretical study performed throughout Chapters 3 and 4 is used to
optimize the performance of a one-way real world test bed.
Chapter 5 studies the n-hop multiple-input multiple-output (MIMO) AF relay network.
Novel techniques are developed using random dynamical system (RDS) theory
and Lyapunov exponents to establish capacity and power scaling laws for the network
as n grows large. One of the main conclusions is that the average transmit
power must grow at an exponential rate if capacity decay across the network is to
be avoided.
Chapter 6 constitutes the final research chapter. In it, the techniques used to study
peak-power constrained OFDM-based networks are combined with those developed
in Chapter 5, which were used to study capacity and power scaling for multihop
AF networks. The conclusion of this is that incorporating OFDM into peak-power
constrained multihop AF relay networks will cause the capacity along each of the
network's eigenchannels to decay exponentially. Finally, we show that the effects of
distortion can be circumvented by ensuring the number of antennas at each node
scales at a super-linear rate with the number of hops within the network.</p
Information-theoretic analysis of the complexity and compression of network topologies
Get PDFNetworks around us are growing larger than ever. The rapid growth of communication networks can clearly be seen in the emergence of 5G and the movement towards 6G, in which the connection density can be as high as ten million devices per square kilometer. This growth is not limited to communication networks and can be seen in other types of networks such as social networks etc. Graph theory has always been the primary mathematical tool for simulating networks. Traditional ways of storing graphs are known to be very complex, even for small graphs. When faced with graphs with millions of nodes, we are inevitably faced with questions about the information content inherent in their topology. We need to answer questions such as how complex these giant graphs are, and how much we are able to compress them. To answer questions like this, we use the help of information theory. Information theory is one of the most powerful tools to analyse the information content of random variables. Treating network graphs as random variables, we start by categorizing them into three main groups based on their behaviour in real-life scenarios. While Shannon's entropy is studied for classic random graph generators, we use Kolmogorov complexity to study the complexity of static networks, and entropy rate to model and study dynamic networks. After studying the complexity of graphs, we move on to study their compression to their theoretical limit. Because of their vast application in different domains, we mainly study optimal tree compression. Finally, the application of our findings to two major engineering problems is highlighted: network routing, and graph neural networks. It is shown how both of these fields can benefit from graph compression and an information-theoretic analysis of the complexity of graphs
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Modeling and signal processing for flash memory
Get PDFThis thesis examines the effects of noise and interference on the performance of NAND flash memory.
Chapter 3 studies the probabilistic input/output relation between the data stored and the read threshold voltage of a cell and generalizes it to a group of cells. It is then concluded that adjacent cells are correlated due to common aggressors. This motivates the study of adequate signal processing techniques to optimize the reliability performance.
Chapter 4 proposes two techniques that can reduce the error rate in light of the result of the previous chapter. The first is based on dividing a group of cells into sub-groups and detecting each sub-group independently. The second approximates the flash system model by a hidden Markov model, then uses the sum-product algorithm to detect the inputs. The soft outputs of the proposed detectors are passed on to the ECC soft decoder. It is shown that the second approach provides significant improvements. Then, it is shown that quantization negatively affects the performance of the sum-product algorithm more in comparison with the first approach. To partially mitigate this effect, an iterative detection/decoding strategy is proposed and shown to improve the performance.
Chapter 5 proposes a novel data representation scheme that provides a trade-off between reliability and the amount of data stored per cell, and partially mitigates the effects of device degradation. The scheme divides the stored data into two streams stored in the indices and levels of the non-erased cells, respectively, allowing the first to be detected without any knowledge about the channel. The simulation results show an improvement in the error rate while partially mitigating the need to track the channel parameters and the read references as the device degrade.</p
Modelling large scale inhomogeneous multihop wireless networks
Get PDFSociety’s burgeoning demand for ubiquitous wireless connectivity, the drive for
greater efficiency and profitability in industry and agriculture sectors and the
growth in e-Health, for example, have led to increased densification, cohabitation
of disparate technologies and increased inhomogeneity in wireless networks.
Whereas this densification enables the deployment of multihop relaying, which
yields a number of tangible benefits, the performance analysis of large-scale relay
networks (≫ 2 hops) is non-trivial and the relay selection for optimal performance
can introduce a high processing complexity, typically high order polynomial time;
potentially unsuited to lower power and edge-processing technologies.
This thesis is concerned with the modelling and analysis of large-scale in
homogeneous multihop wireless networks in a number of contexts and ‘real
world’ application scenarios. We analyse 2-hop connectivity within geometrically
bounded networks, deriving a closed form expression for connection probability,
and apply it in a device-to-device (D2D) context within a cellular-type network.
We show that appreciable reductions in the burden on the cellular-infrastructure
can be realised with this D2D approach.
Progressing to larger-scale networks, we propose a novel approach to multihop
route optimisation by considering the limit of infinite relay node density, yielding
an optimised equivalent continuous relay path, or continuum. The model is
carefully constructed to maintain a constant connection density even though the
node density scales without bound. This provides a formulation for determining
performance extrema using methods from the field of calculus of variations.
With our model, we show that processing complexity scales linearly with the
number of points that sample the continuous path, which can be lower than the
number of relay nodes in a large scale network. We demonstrate the effectiveness
of this new approach and its potential by considering a network subjected to
point sources of interference and to the problem of optimal covert routing in
the presence of an eavesdropper
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Identification and mitigation of attacks in trust-based distributed communication networks
Get PDFNetwork security is a vital component of modern wireless communications, particularly as a result of the value of data being shared within those networks. Trust inference has provided an intuitive and accessible way for distributed networks to protect themselves in lieu of a central authority, such as a base-station. However, the trust metrics themselves can be manipulated by malicious parties in order to disrupt the network.
This thesis examines the vulnerabilities that exist within distributed communication networks that rely on trust inference for security. Chapter 3 studies the landscape of trust network vulnerabilities and presents a design for a unified framework of trust network attacks from an observer's perspective, based on the comparison and collation of pre-attack symptoms. This is then used as the basis for a novel contextual characterisation method for the classification of nodes' behaviour into possible attack scenarios, which is implemented using a support vector machine.
Chapter 4 presents the design and analysis of a trust-based data management and aggregation protocol for facilitating secure self-management of distributed network nodes. This is accompanied by the proposal for a trust-based data aggregation protocol to support network functionality, which enables coexistence with malicious nodes in the network by using their input to reinforce network decisions.
In Chapter 5, we devise a comprehensive model for node behaviour using a hidden Markov model approach, and show that the Baum Welch algorithm can be used to estimate the transition and emission probabilities of a node. Subsequently, it is shown that the use of a long short-term memory RNN facilitates the early classification and extrapolation of the node's behaviour, such that its attack probability can be estimated to a high degree of accuracy even before sufficient observations are collected
- …
