Fabian Seguel

Enhancement of the QoS in an OFDMA/VLC system

A resource allocation optimization for indoor Optical Wireless Communications is proposed. A smart coordinator tries to maximize the channel capacity by enabling or disabling the communication of the devices subject to system constraints. Binary Cuckoo Search and Genetic Algorithms were used in order to solve the binary NP-HARD optimization problem. The results show that the total channel capacity slowly decrease while the number of users increase. On the other hand, per-user capacity decrease faster than the total capacity. The minimum QoS of the system is achieved. Binary Cuckoo Search overcomes Genetic Algorithms when dealing with the proposed optimization problem.

A new VLC system for localization in underground mining tunnels

In this paper we present a new system that uses new technology that has the ability to transmit data using light emitting diodes (LED) and a photodetector (PD) capable of sensing very soft changes of light. The system developed in conjunction with the trilateration technique can be used for localizing personnel as well as machinery in underground mining tunnels. Simulations are performed, and comparisons are made with other well-known techniques in terms of the location estimation error. Overall, it is shown that the proposed system is a viable option for localization in underground mining tunnels.

Optimal Resource Allocation for Non-Real Time Visible Light Communication Networks

In this paper, we consider the optimal joint resource allocation problem of subcarrier and power for non-real time wireless networks that use visible light communication (VLC) technology. VLC has been recognized as a promising technology as it allows to transmit data in considerably higher orders of magnitude compared to traditional radio frequency (RF) methods [3]. Therefore, it is expected that most of the current existing protocols for resource allocation using traditional RF technologies will adapt to VLC technology in the near future. We model the resource allocation problem by means of a mixed integer nonlinear optimization problem that we further linearize using a piecewise linear approximation method. The latter allows to compute optimal and near optimal solutions for the problem. Optimal solutions can be obtained as long as the number of line segments is sufficiently large. Subsequently, we propose a variable neighborhood search based decomposition procedure that allows to compute, in average, tight near optimal solutions in less than one second compared to the high CPU time required by the piecewise linear model.

A Meta-heuristic Approach for Copper Price Forecasting

The price of copper and its variations represent a very important financial issue for mining companies and for the Chilean government because of its impact on the national economy. The price of commodities such as copper is highly volatile, dynamic and troublous. Due to this, forecasting is very complex. Using publicly data from October 24th of 2013 to August 29th of 2014 a multivaried based model using meta-heuristic optimization techniques is proposed. In particular, we use Genetic Algorithms and Simulated Annealing in order to find the best fitting parameters to forecast the variation on the copper price. A non-parametric test proposed by Timmermann and Pesaran is used to demonstrate the forecasting capacity of the models. Our numerical results show that the Genetic Algorithmic approach has a better performance than Simulated Annealing, being more effective for long range forecasting.

Finding Degree Constrained k-Cardinality Minimum Spanning Trees for Wireless Sensor Networks

In this paper, we consider the degree constrained k-cardinality minimum spanning tree network problem (k-DCMST). This problem arises as a combination of two classical optimization problems, namely the degree constrained and k-minimum spanning tree problems (Resp. DCMST and k-MST). Let G(V, E) be a connected undirected graph formed with vertex and edge sets V and E, respectively. The DCMST problem asks for a minimum spanning tree where each maximum vertex degree is limited to a certain constant d lower than the cardinality of V minus one whilst the k-MST asks for a minimum spanning sub-tree formed with k nodes chosen from set V. Consequently, the k-DCMST asks for a sub-tree formed with k vertices where each vertex has degree lower than or equal to d. This problem is mainly motivated from the domain of wireless sensor networks where connected backbone sub-tree topologies will be mandatorily required for future technologies in order to connect any network under the internet of things paradigm. Vertex degree constraints arise naturally in order to avoid overloaded nodes in the network. We propose two compact formulations for this problem. More precisely, a Miller-Tucker-Zemlin constrained version and a single flow based formulation that we further strengthen by using the Handshaking lemma and with valid inequalities adapted from the DCMST and dominating tree problems. Numerical results are given for complete and disk graph instances for different degree values. Our preliminary numerical results indicate that the flow based model allows one to obtain optimal solutions in less CPU time for most of the instances.

Evaluation of Llaima volcano activities for localization and classification of LP, VT and TR events

Abstract Evaluation of seismic signals is one of the most important research topics on Volcanology. Volcanoes have daily activity; therefore, high speed evaluation of recorded signals is a challenge for improving the study of the natural phenomena occurring inside these natural formations. The aim of this paper is the evaluation (denoising, localization and classification) and analysis of Llaima volcano activities, one of the most actives volcanoes in South America. Different already proposed methods, such as, Butterworth, Spectral Subtraction (SS) and Wiener Filter (WF) are compared to the proposed Modified Spectral Subtraction (MSS) and Modified Wiener Filter (MWF) to find the best method for denoising the volcano signals. Then, event localization based on received signals of volcano is performed. In this step, Time Delay Estimation (TDE)-based method is used on data acquired from 3 mechanical sensors located in the volcano area. The proposed method is used to estimate the area for event location. The proposed denoising methods make the starting point for the event more evident to increase the localization accuracy for events where the starting point is difficult to find. In the last step, a method based on the novel DNN technique is proposed to classify the three main events occurring in the Llaima volcano (TR (Tremor), LP (Long Period) and VT (Volcano Tectonic)).

New Formulations for an Optimal Connectivity Approach for Mobile Ad-hoc Networks

In this paper, we propose new formulations for the optimal connectivity of a tree backbone topology for mobile ad-hoc networks (MANETs). Applications of MANETs include military communications, emergency and disaster recovery, and e-commerce to name a few. Formally, given a graph \(G=(V \cup K,E)\) with set of wireless sensor nodes V, a set of connection links E, and a set of K users, the problem is to find a backbone spanning tree network topology with as many leaves as possible in order to maximize capacity at the lowest power costs for the users which are connected to the leaf nodes of the backbone network. For this purpose, we model a MANET by means of disk graphs where each disk represents the Euclidean distance transmission range of a node \(v \in V\). We propose an exponential and a compact polynomial formulation for the problem. The exponential model is characterized with constraints from the classical maximum leaf spanning tree polytope [18] whilst the compact formulation is characterized with constraints adapted from the classical minimum dominating tree problem [2]. The latter formulation is further strengthened with selected valid inequalities referred to as generalized sub-tour elimination constraints [9]. Our preliminary numerical results indicate that the compact model with additional valid inequalities allows to solve instances to optimality in significantly short CPU time for transmission distances ranging from 100 to 150 ms.

Proposed energy based method for light receiver localization in underground mining

Light source localization is one of the most functional area in light signal processing. Various methods are used for light localization. Most of these methods are based on image signal processing on 2-D recorded image of environment but they need more devices for image recording. The aim of this paper, is localization of a photodetector or light receiver. These light receivers are used in connection with light sources, because as seen, light is not emitted from these photodetectors. Then, actually, the image processing based methods are useless in this application. The method in this paper is based on energy calculation in different frequency range based on received light from different light sources. Firstly, the effect of different light sources is separated by smart filtering. Then, the place of light receiver is estimated by energy calculation on specific frequency range and finding the intersection of spheres due to each light source. The benefit of the proposed method is using 1-D signal and needless of any specific device and in all of environment conditions. This proposed method can be implemented in real time due to low computational complexity. Also, this method has less complexity in compare with image processing based methods. Moreover, the accuracy of the proposed method is equal or sometimes better than the image processing based methods. The proposed method in this paper has been evaluated in different environmental conditions and the results shows a good accuracy. Also, this method can be used in underground mining applications.