Multicast traffic filtering for sampled value process bus networks

Ethernet is a key component of the standards used for digital process buses in transmission substations, namely IEC 61850 and IEEE Std 1588-2008 (PTPv2). These standards use multicast Ethernet frames that can be processed by more than one device. This presents some significant engineering challenges when implementing a sampled value process bus due to the large amount of network traffic. A system of network traffic segregation using a combination of Virtual LAN (VLAN) and multicast address filtering using managed Ethernet switches is presented. This includes VLAN prioritisation of traffic classes such as the IEC 61850 protocols GOOSE, MMS and sampled values (SV), and other protocols like PTPv2. Multicast address filtering is used to limit SV/GOOSE traffic to defined subsets of subscribers. A method to map substation plant reference designations to multicast address ranges is proposed that enables engineers to determine the type of traffic and location of the source by inspecting the destination address. This method and the proposed filtering strategy simplifies future changes to the prioritisation of network traffic, and is applicable to both process bus and station bus applications.

Defective interfering viral particles in acute dengue infections

While much of the genetic variation in RNA viruses arises because of the error-prone nature of their RNA-dependent RNA polymerases, much larger changes may occur as a result of recombination. An extreme example of genetic change is found in defective interfering (DI) viral particles, where large sections of the genome of a parental virus have been deleted and the residual sub-genome fragment is replicated by complementation by co-infecting functional viruses. While most reports of DI particles have referred to studies in vitro, there is some evidence for the presence of DI particles in chronic viral infections in vivo. In this study, short fragments of dengue virus (DENV) RNA containing only key regulatory elements at the 3' and 5' ends of the genome were recovered from the sera of patients infected with any of the four DENV serotypes. Identical RNA fragments were detected in the supernatant from cultures of Aedes mosquito cells that were infected by the addition of sera from dengue patients, suggesting that the sub-genomic RNA might be transmitted between human and mosquito hosts in defective interfering (DI) viral particles. In vitro transcribed sub-genomic RNA corresponding to that detected in vivo could be packaged in virus like particles in the presence of wild type virus and transmitted for at least three passages in cell culture. DENV preparations enriched for these putative DI particles reduced the yield of wild type dengue virus following co-infections of C6-36 cells. This is the first report of DI particles in an acute arboviral infection in nature. The internal genomic deletions described here are the most extensive defects observed in DENV and may be part of a much broader disease attenuating process that is mediated by defective viruses.

Identification of a cryptic Prokaryotic promoter within the cDNA encoding the 5′ End of Dengue Virus RNA Genome

Infectious cDNA clones of RNA viruses are important research tools, but flavivirus cDNA clones have proven difficult to assemble and propagate in bacteria. This has been attributed to genetic instability and/or host cell toxicity, however the mechanism leading to these difficulties has not been fully elucidated. Here we identify and characterize an efficient cryptic bacterial promoter in the cDNA encoding the dengue virus (DENV) 5′ UTR. Following cryptic transcription in E. coli, protein expression initiated at a conserved in-frame AUG that is downstream from the authentic DENV initiation codon, yielding a DENV polyprotein fragment that was truncated at the N-terminus. A more complete understanding of constitutive viral protein expression in E. coli might help explain the cloning and propagation difficulties generally observed with flavivirus cDNA.

The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the network

Sample complexity results from computational learning theory, when applied to neural network learning for pattern classification problems, suggest that for good generalization performance the number of training examples should grow at least linearly with the number of adjustable parameters in the network. Results in this paper show that if a large neural network is used for a pattern classification problem and the learning algorithm finds a network with small weights that has small squared error on the training patterns, then the generalization performance depends on the size of the weights rather than the number of weights. For example, consider a two-layer feedforward network of sigmoid units, in which the sum of the magnitudes of the weights associated with each unit is bounded by A and the input dimension is n. We show that the misclassification probability is no more than a certain error estimate (that is related to squared error on the training set) plus A3 √((log n)/m) (ignoring log A and log m factors), where m is the number of training patterns. This may explain the generalization performance of neural networks, particularly when the number of training examples is considerably smaller than the number of weights. It also supports heuristics (such as weight decay and early stopping) that attempt to keep the weights small during training. The proof techniques appear to be useful for the analysis of other pattern classifiers: when the input domain is a totally bounded metric space, we use the same approach to give upper bounds on misclassification probability for classifiers with decision boundaries that are far from the training examples.

Exponentiated gradient algorithms for conditional random fields and max-margin Markov Networks

Log-linear and maximum-margin models are two commonly-used methods in supervised machine learning, and are frequently used in structured prediction problems. Efficient learning of parameters in these models is therefore an important problem, and becomes a key factor when learning from very large data sets. This paper describes exponentiated gradient (EG) algorithms for training such models, where EG updates are applied to the convex dual of either the log-linear or max-margin objective function; the dual in both the log-linear and max-margin cases corresponds to minimizing a convex function with simplex constraints. We study both batch and online variants of the algorithm, and provide rates of convergence for both cases. In the max-margin case, O(1/ε) EG updates are required to reach a given accuracy ε in the dual; in contrast, for log-linear models only O(log(1/ε)) updates are required. For both the max-margin and log-linear cases, our bounds suggest that the online EG algorithm requires a factor of n less computation to reach a desired accuracy than the batch EG algorithm, where n is the number of training examples. Our experiments confirm that the online algorithms are much faster than the batch algorithms in practice. We describe how the EG updates factor in a convenient way for structured prediction problems, allowing the algorithms to be efficiently applied to problems such as sequence learning or natural language parsing. We perform extensive evaluation of the algorithms, comparing them to L-BFGS and stochastic gradient descent for log-linear models, and to SVM-Struct for max-margin models. The algorithms are applied to a multi-class problem as well as to a more complex large-scale parsing task. In all these settings, the EG algorithms presented here outperform the other methods.

Molecular classification of cutaneous malignant melanoma by gene expression profiling

The most common human cancers are malignant neoplasms of the skin. Incidence of cutaneous melanoma is rising especially steeply, with minimal progress in non-surgical treatment of advanced disease. Despite significant effort to identify independent predictors of melanoma outcome, no accepted histopathological, molecular or immunohistochemical marker defines subsets of this neoplasm. Accordingly, though melanoma is thought to present with different 'taxonomic' forms, these are considered part of a continuous spectrum rather than discrete entities. Here we report the discovery of a subset of melanomas identified by mathematical analysis of gene expression in a series of samples. Remarkably, many genes underlying the classification of this subset are differentially regulated in invasive melanomas that form primitive tubular networks in vitro, a feature of some highly aggressive metastatic melanomas. Global transcript analysis can identify unrecognized subtypes of cutaneous melanoma and predict experimentally verifiable phenotypic characteristics that may be of importance to disease progression.

Emergent task-specific object semantics through distributed experience networks

Autonomous development of sensorimotor coordination enables a robot to adapt and change its action choices to interact with the world throughout its lifetime. The Experience Network is a structure that rapidly learns coordination between visual and haptic inputs and motor action. This paper presents methods which handle the high dimensionality of the network state-space which occurs due to the simultaneous detection of multiple sensory features. The methods provide no significant increase in the complexity of the underlying representations and also allow emergent, task-specific, semantic information to inform action selection. Experimental results show rapid learning in a real robot, beginning with no sensorimotor mappings, to a mobile robot capable of wall avoidance and target acquisition.

Aerial image analysis using spiking neural networks with application to power line corridor monitoring

Trees, shrubs and other vegetation are of continued importance to the environment and our daily life. They provide shade around our roads and houses, offer a habitat for birds and wildlife, and absorb air pollutants. However, vegetation touching power lines is a risk to public safety and the environment, and one of the main causes of power supply problems. Vegetation management, which includes tree trimming and vegetation control, is a significant cost component of the maintenance of electrical infrastructure. For example, Ergon Energy, the Australia’s largest geographic footprint energy distributor, currently spends over $80 million a year inspecting and managing vegetation that encroach on power line assets. Currently, most vegetation management programs for distribution systems are calendar-based ground patrol. However, calendar-based inspection by linesman is labour-intensive, time consuming and expensive. It also results in some zones being trimmed more frequently than needed and others not cut often enough. Moreover, it’s seldom practicable to measure all the plants around power line corridors by field methods. Remote sensing data captured from airborne sensors has great potential in assisting vegetation management in power line corridors. This thesis presented a comprehensive study on using spiking neural networks in a specific image analysis application: power line corridor monitoring. Theoretically, the thesis focuses on a biologically inspired spiking cortical model: pulse coupled neural network (PCNN). The original PCNN model was simplified in order to better analyze the pulse dynamics and control the performance. Some new and effective algorithms were developed based on the proposed spiking cortical model for object detection, image segmentation and invariant feature extraction. The developed algorithms were evaluated in a number of experiments using real image data collected from our flight trails. The experimental results demonstrated the effectiveness and advantages of spiking neural networks in image processing tasks. Operationally, the knowledge gained from this research project offers a good reference to our industry partner (i.e. Ergon Energy) and other energy utilities who wants to improve their vegetation management activities. The novel approaches described in this thesis showed the potential of using the cutting edge sensor technologies and intelligent computing techniques in improve power line corridor monitoring. The lessons learnt from this project are also expected to increase the confidence of energy companies to move from traditional vegetation management strategy to a more automated, accurate and cost-effective solution using aerial remote sensing techniques.

Parallel implementation of stochastic simulation for large scale cellular processes

Experimental and theoretical studies have shown the importance of stochastic processes in genetic regulatory networks and cellular processes. Cellular networks and genetic circuits often involve small numbers of key proteins such as transcriptional factors and signaling proteins. In recent years stochastic models have been used successfully for studying noise in biological pathways, and stochastic modelling of biological systems has become a very important research field in computational biology. One of the challenge problems in this field is the reduction of the huge computing time in stochastic simulations. Based on the system of the mitogen-activated protein kinase cascade that is activated by epidermal growth factor, this work give a parallel implementation by using OpenMP and parallelism across the simulation. Special attention is paid to the independence of the generated random numbers in parallel computing, that is a key criterion for the success of stochastic simulations. Numerical results indicate that parallel computers can be used as an efficient tool for simulating the dynamics of large-scale genetic regulatory networks and cellular processes

An efficient public key management regime for vehicular ad hoc networks (VANETS)

The primary goal of the Vehicular Ad Hoc Network (VANET) is to provide real-time safety-related messages to motorists to enhance road safety. Accessing and disseminating safety-related information through the use of wireless communications technology in VANETs should be secured, as motorists may make critical decisions in dealing with an emergency situation based on the received information. If security concerns are not addressed in developing VANET systems, an adversary can tamper with, or suppress, the unprotected message to mislead motorists to cause traffic accidents and hazards. Current research on secure messaging in VANETs focuses on employing the certificate-based Public Key Infrastructure (PKI) scheme to support message encryption and digital signing. The security overhead of such a scheme, however, creates a transmission delay and introduces a time-consuming verification process to VANET communications. This thesis has proposed a novel public key verification and management approach for VANETs; namely, the Public Key Registry (PKR) regime. Compared to the VANET PKI scheme, this new approach can satisfy necessary security requirements with improved performance and scalability, and at a lower cost by reducing the security overheads of message transmission and eliminating digital certificate deployment and maintenance issues. The proposed PKR regime consists of the required infrastructure components, rules for public key management and verification, and a set of interactions and associated behaviours to meet these rule requirements. This is achieved through a system design as a logic process model with functional specifications. The PKR regime can be used as development guidelines for conforming implementations. An analysis and evaluation of the proposed PKR regime includes security features assessment, analysis of the security overhead of message transmission, transmission latency, processing latency, and scalability of the proposed PKR regime. Compared to certificate-based PKI approaches, the proposed PKR regime can maintain the necessary security requirements, significantly reduce the security overhead by approximately 70%, and improve the performance by 98%. Meanwhile, the result of the scalability evaluation shows that the latency of employing the proposed PKR regime stays much lower at approximately 15 milliseconds, whether operating in a huge or small environment. It is therefore believed that this research will create a new dimension to the provision of secure messaging services in VANETs.

Frequency response function based structural damage detection using artificial neural networks

Damage detection in structures has become increasingly important in recent years. While a number of damage detection and localization methods have been proposed, few attempts have been made to explore the structure damage with frequency response functions (FRFs). This paper illustrates the damage identification and condition assessment of a beam structure using a new frequency response functions (FRFs) based damage index and Artificial Neural Networks (ANNs). In practice, usage of all available FRF data as an input to artificial neural networks makes the training and convergence impossible. Therefore one of the data reduction techniques Principal Component Analysis (PCA) is introduced in the algorithm. In the proposed procedure, a large set of FRFs are divided into sub-sets in order to find the damage indices for different frequency points of different damage scenarios. The basic idea of this method is to establish features of damaged structure using FRFs from different measurement points of different sub-sets of intact structure. Then using these features, damage indices of different damage cases of the structure are identified after reconstructing of available FRF data using PCA. The obtained damage indices corresponding to different damage locations and severities are introduced as input variable to developed artificial neural networks. Finally, the effectiveness of the proposed method is illustrated and validated by using the finite element modal of a beam structure. The illustrated results show that the PCA based damage index is suitable and effective for structural damage detection and condition assessment of building structures.