Reactive Robust Routing: Anomaly Localization and Routing Reconfiguration for Dynamic Networks

This paper presents a novel approach to deal with dynamic and highly uncertain traffic in dynamic network scenarios. The Reactive Robust Routing (RRR) approach is introduced, a combination of proactive and reactive techniques to improve network efficiency and robustness, simplifying network operation. RRR optimizes routing for normal-operation traffic, using a time-varying extension of the already established Robust Routing technique that outperforms the stable approach. To deal with anomalous and unexpected traffic variations, RRR uses a fast anomaly detection and localization algorithm that rapidly detects and localizes abrupt changes in traffic flows, permitting an accurate routing adaptation. This algorithm presents well-established optimality properties in terms of detection/localization rates and localization delay, which allows for generalization of results, independently of particular evaluations. The algorithm is based on a novel parsimonious model for traffic demands which allows for detection of anomalies using easily available aggregated-traffic measurements, reducing the overheads of data collection.     

Leveraging temporal and spatial separations with the 24-hour knowledge factory paradigm

The 24-H Knowledge Factory facilitates collaboration between geographically and temporally distributed teams. The teams themselves form a strategic partnership whose joint efforts contribute to the completion of a project. Project-related tasks are likewise distributed, allowing tasks to be completed on a continuous basis, regardless of the constraints of any one team’s working hours. However, distributing a single task between multiple teams necessitates a handoff process, where one team’s development efforts and task planning are communicated from one team ending their shift to the next that will continue the effort. Data management is, therefore, critical to the success of this business model. Efficiency in data management is achieved through a strategic leveraging of key tools, models, and concepts.     

Explanation and reliability of prediction models: the case of breast cancer recurrence

In this paper, we describe the first practical application of two methods, which bridge the gap between the non-expert user and machine learning models. The first is a method for explaining classifiers’ predictions, which provides the user with additional information about the decision-making process of a classifier. The second is a reliability estimation methodology for regression predictions, which helps the users to decide to what extent to trust a particular prediction. Both methods are successfully applied to a novel breast cancer recurrence prediction data set and the results are evaluated by expert oncologists.     

Predictive functional control based on an adaptive fuzzy model of a hybrid semi-batch reactor

In this paper a new approach to the control of a nonlinear, time-varying process is proposed. It is based on a recursive version of the fuzzy identification method and predictive functional control. First, the recursive fuzzy identification method is derived, after which it is used in connection with fuzzy predictive functional control to construct an adaptive fuzzy predictive functional controller. The adaptive FPFC is then tested on a nonlinear, time-varying, semi-batch reactor process and compared with the standard FPFC, which uses non-adaptive fuzzy model. The simulation results are promising; they indicate that the control of time-varying, nonlinear processes with the FPFC can be improved with the use of an adaptive fuzzy model. An improvement in reference tracking and disturbance rejection can be observed, but the main advantage is the reduced number of switchings between hot and cold water. This is an important improvement in the case of real applications.     

Low temperature co-fired ceramic (LTCC)-based biosensor for continuous glucose monitoring

This article presents design, fabrication and testing of a miniature ceramic-based biosensor which is destined for continuous glucose monitoring. It is fabricated using well known LTCC (low temperature co-fired ceramics) technology. The biosensor consists of a microreaction chamber, three thick-film electrodes and a microdialysis tube. The detection process is based on oxidation of glucose by molecular oxygen in the presence of enzyme - glucose oxidase, GOx. One of the reaction products is a hydrogen peroxide which is detected amperometrically during its oxidation at a working electrode. Computational fluid dynamics (CFD) analysis was used for optimization of the biosensor geometry and considerations of the influence of operating conditions (flow rate) on mass transfer (diffusion) of the glucose within the microreaction chamber. Tests of the developed LTCC-based biosensor indicate its linear response to glucose concentration up to 9 mM with a relatively high sensitivity of about 147 nA/mM. Moreover, the properties of the presented ceramic biosensor are compared with properties of a similar device made in silicon/glass and in Perspex^®.     

Consistent Sets of Secondary Structures in Proteins

Ab initio predictions of secondary structures in proteins have to combine local predictions, based on short fragments of the protein sequence, with consistency restrictions, as not all locally plausible predictions may be simultaneously true. We use the fact that secondary structures are patterns of hydrogen bonds and that a single residue can participate in hydrogen bonds of at most one secondary structure. Consistency of fixed-sized pieces of secondary structures is the easiest to approximate and we formalize it as 1-2 matching problem. Consistency of entire secondary structures is a version of set packing. We also investigate how to form a simple problem if we add the requirement that the secondary structure and the loops that connect them fit together in a metric space. Every problem that we investigated is MAX-SNP hard and it has a constant factor approximation. Computational experience suggests that in biological instances, we can find nearly optimal solutions using heuristics.     

Self-optimizing generalized adaptive notch filters — comparison of three optimization strategies

The paper provides comparison of three different approaches to on-line tuning of generalized adaptive notch filters (GANFs) — the algorithms used for identification/tracking of quasi-periodically varying dynamic systems. Tuning is needed to adjust adaptation gains, which control tracking performance of GANF algorithms, to the unknown and/or time time-varying rate of system nonstationarity. Two out of three compared approaches are classical solutions — the first one incorporates sequential optimization of adaptation gains while the second one is based on the concept of parallel estimation. The main contribution of the paper is that it suggests the third way — it shows that the best results can be achieved when both approaches mentioned above are combined in a judicious way. Such joint sequential/parallel optimization preserves advantages of both treatments: adaptiveness (sequential approach) and robustness to abrupt changes (parallel approach). Additionally the paper shows how, using the concept of surrogate outputs, one can extend the proposed single-frequency algorithm to the multiple frequencies case, without falling into the complexity trap known as the “curse of dimensionality”.     

Oscillations and transfer properties of relay feedback systems with time-delay linear plants

Oscillations and transfer properties of relay feedback systems with linear plants having a dead time (time delay) are considered. Poincare maps are found explicitly in terms of the system parameters. The oscillatory and transfer properties are presented in the format of the locus of a perturbed relay system (LPRS), which represents the continuum of all possible periodic solutions and transfer properties of the system with a given linear plant. Conditions of local orbital asymptotic stability are found in terms of the system parameters. The proposed approach is illustrated by examples.     

Accelerating scientific computations with mixed precision algorithms

On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution. The approach presented here can apply not only to conventional processors but also to other technologies such as Field Programmable Gate Arrays (FPGA), Graphical Processing Units (GPU), and the STI Cell BE processor. Results on modern processor architectures and the STI Cell BE are presented.

Program summary

Program title: ITER-REFCatalogue identifier: AECO_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AECO_v1_0.htmlProgram obtainable from: CPC Program Library, Queen's University, Belfast, N. IrelandLicensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.htmlNo. of lines in distributed program, including test data, etc.: 7211No. of bytes in distributed program, including test data, etc.: 41 862Distribution format: tar.gzProgramming language: FORTRAN 77Computer: desktop, serverOperating system: Unix/LinuxRAM: 512 MbytesClassification: 4.8External routines: BLAS (optional)Nature of problem: On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution.Solution method: Mixed precision algorithms stem from the observation that, in many cases, a single precision solution of a problem can be refined to the point where double precision accuracy is achieved. A common approach to the solution of linear systems, either dense or sparse, is to perform the LU factorization of the coefficient matrix using Gaussian elimination. First, the coefficient matrix A is factored into the product of a lower triangular matrix L and an upper triangular matrix U. Partial row pivoting is in general used to improve numerical stability resulting in a factorization PA=LU, where P is a permutation matrix. The solution for the system is achieved by first solving Ly=Pb (forward substitution) and then solving Ux=y (backward substitution). Due to round-off errors, the computed solution, x, carries a numerical error magnified by the condition number of the coefficient matrix A. In order to improve the computed solution, an iterative process can be applied, which produces a correction to the computed solution at each iteration, which then yields the method that is commonly known as the iterative refinement algorithm. Provided that the system is not too ill-conditioned, the algorithm produces a solution correct to the working precision.Running time: seconds/minutes