William Giozza

The choice of the best among the shortest routes in transparent optical networks

This work introduces the problem of the best choice among M combinations of the shortest paths for dynamic provisioning of lightpaths in all-optical networks. To solve this problem in an optimized way (shortest path and load balance), a new fixed routing algorithm, named Best among the Shortest Routes (BSR), is proposed. The BSRs performance is compared in terms of blocking probability and network utilization with Dijkstras shortest path algorithm and others algorithms proposed in the literature. The evaluated scenarios include several representative topologies for all-optical networking and different wavelength conversion architectures. For all studied scenarios, BSR achieved superior performance.

Hybrid Algorithms for Routing and Assignment Wavelengths in Optical Networks

This paper presents a strategy for the solution of the WDM optical networks planning. Specifically, the problem of Routing and Wavelength Allocation (RWA) in order to minimize the amount of wavelengths used. In this case, the problem is known as the Min-RWA. Two meta-heuristics (Tabu Search and Simulated Annealing) are applied to take solutions of good quality and high performance. The key point is the degradation of the maximum load on the virtual links in favor of minimization of number of wavelengths used; the objective is to find a good compromise between the metrics of virtual topology (load in Gb/s) and of the physical topology (quantity of wavelengths). The simulations suggest good results when compared to some existing in the literature.

Wavelength conversion architectures in OBS networks

Most studies about Optical Burst Switched (OBS) networks have assumed availability of full wavelength conversion throughout the network. In an optical network with full wavelength conversion capability, every output on each node is provided with a wavelength converter, resulting in the lowest bound for blocking probability. However, full wavelength conversion capability requires a large number of wavelength converters, which is, and will remain, quite expensive in the foreseeable future. Therefore, it is expected that conversion capability will be limited and only sparsely deployed. In this work we apply to OBS networks the main architectures for wavelength conversion proposed for circuit switched networks. These architectures operate with limited wavelength conversion capability. We also investigate the cost benefit and the gain factor related to full complete wavelength conversion for each architecture. In addition, we propose a new methodology to define the number of wavelength converters in order to achieve the best cost benefit, regardless of the traffic load.

Classification Strategy to Mitigate Unfairness in All-Optical Networks

This paper investigates the fairness problem in all-optical networks without wavelength conversion under dynamic traffic loads. We propose a new classification strategy to mitigate the fairness problem called Fair-Fit. Our strategy is based on previous simulation studies that are carried out off-line. In addition, we introduce an equation that allows one to compare the source-destination pairs with the best and the worst blocking probability. Our simulation results show that Fair-Fit may result in better fairness results than earlier strategies proposed in the literature.

A new approach for performance evaluation of survivability services in all-optical networks

We propose in this paper a new set of measures (blocking probability, restorability, blocking probability per source-destination pair, blocking probability due to the absence of resources in the backup route, etc.) for performance evaluation of survivability services in all-optical networks. A comparison between the Two Step with Backtracking (TSB) and the Simple Restoration (SR), two important references of all-optical network survivability schemes is made using such set of measures. We also propose a new technique to provide survivability for all-optical networks, named Adaptive Survivability (ADPS). Besides being more flexible, we show that our scheme has a more balanced performance in terms of blocking probability and restorability when compared to the Two Step with Backtracking and the Simple Restoration techniques.

Cost Benefit Study of Sparse-PartialWavelength Conversion Architectures

In this paper, we propose a new wavelength converter placement scheme called Adaptive First Load Priority, useful for designing and planning of optical networks with sparse-partial wavelength conversion architecture. The adaptive FLP is an extension of the FLP (First Load Priority) placement scheme proposed by authors in a previous work [13]. Initially, we provide a performance evaluation of FLP. Comparing to others wavelength converter placement schemes proposed in the literature like XC scheme [5], our results show that FLP achieves the best performance at all scenarios studied, very close to that one achieved with a full-complete wavelength conversion architecture. However, we could observe that, after a certain threshold, just adding moreWavelength Converter Routers (WCR) with the FLP scheme may not have enough improvement in network performance that justifies a higherWCR cost. The Adaptive First Load Priority scheme was conceived in order to minimize the amount of WCRs used by FLP scheme, improving its cost benefit. A cost-benefit (ratio gain) study showing a superior performance of Adaptive FLP scheme is presented.

Increasing the Dependability of IoT Middleware with Cloud Computing and Microservices

Internet of Things (IoT) has left the experimental field and is reaching the final consumer in areas such as residential automation, health, transportation and government support. Since these are applications intrinsically linked to the physical world, they require more attention in aspects related to their dependability. Focusing on its availability and reliability, we present a proposal to apply the Microservices architectural pattern in an IoT middleware with web services in the monolithic architecture. We describe the reengineering that must be done in this middleware and, finally, we analyze the advantages and disadvantages of this approach, highlighting the availability improvement, optimization of the infrastructure resources, the ease of maintenance and evolution, as well as the inclination to the elasticity that the architecture allows.

Which of the shortest paths should we choose? A proposal of routing in the all-optical WDM networks design

We propose to use an iterative heuristic for the shortest path routing in static RWA problem of optical networks. The heuristics performance is compared, in terms of number of wavelengths, with the shortest Dijkstras path algorithm. The results suggest that it is feasible to choose routes, among the shortest paths, minimizing the number of wavelengths. For all studied scenarios, the heuristic achieved superior performance.

Performance Evaluation of Fixed Routing Algorithms in All-Optical Networks under Non-uniform Traffic

Most works about all-optical circuit-switched WDM networks considers a shortest path algorithm for fixed routing, like the Dijkstra’ algorithm. However, recently, two new algorithms, searching for better load balancing and presenting superior performance than Dijkstra´s when submitted to an uniform traffic load, were proposed, the Restricted Routing Technique (RRT) and the Best among the Shortest Routes - (BSR). This work evaluates the performance of these algorithms under a non-uniform traffic load. For each topology studied were observed specific characteristics of the impact of the non-uniform traffic load for each routing algorithm. Overall, performance evaluation results showed that the algorithm BSR achieves for all topologies studied superior performance in terms of blocking probability and fairness.

PCA-Kalman based load forecasting of electric power demand

Electricity demand time series are stochastic processes related to climate, social and economic variables. By predicting the evolution of such time series, electrical load forecasting can be performed in order to support the electrical grid planning. In this paper, we propose a Kalman based load forecasting system for daily demand forecasting. Our proposed approach incorporates a Principal Component Analysis (PCA) of the input variables obtained from linear and nonlinear transformations of the candidate time series. In order to validate our predicting scheme, data collected from Brasília distribution company has been used. Our proposed approach outperforms state-of-the-art approaches based on state space and artificial neural networks.