Hector M. Lugo-Cordero

Particle Swarm Optimization for load balancing in green smart homes

Particle Swarm Optimization (PSO) is a promising evolutionary algorithm, which has been used in a wide range of applications, due to its simple implementation, fast convergence, parallel behavior, and versatility in working with continuous and discrete domains. In this paper, we consider its application to the load balancing problem, in green smart homes. Specifically, an adapted version of the Binary PSO has been used to determine the optimal distribution of energy resources, across different green energy sources in a green smart home. The case study of interest considers the usage of solar and wind energy, as green energy sources for the green smart home. Results demonstrate the effectiveness of the algorithm, in terms of the optimal outcome (efficient distribution of energy resources), finding installation material surplus, and the execution speed of the algorithm.

Evolution of optimal heterogeneous wireless mesh networks

Wireless networks provide feasible solutions for a wide range of applications. Among the applications users may find environment monitoring, Internet access provision, and cellular communications. However, wireless networks have a common problem, independently of the technology used: deploying nodes in such a way, where the resources may be at optimum performance. These resources include map coverage, network power consumption, network cost, node anonymity, and data confidentiality. In this paper, we introduce an algorithm for deploying heterogeneous wireless mesh networks (HMWNs), named Optimal Network Evolver (ONE). ONE utilizes evolutionary computation to obtain the best network configuration for a given map, in order to meet the system requirements. ONE takes into account path loss of free space, trees, and walls. ONE also considers how the frequency of operation of a node is affected by such path loss. At the end of execution, ONE provides HWMNs designers the optimal number of nodes required for the given map, their placement location, and proper configuration. Results show map coverage, network power consumption, node anonymity, and data confidentiality are optimized by ONE on either simple or complex environments.

A novel service-oriented routing algorithm for wireless mesh network

Wireless mesh network (WMN) is a promising technology that can provide cost-effective solutions to cover a rather large area. Despite this important feature, the future generation of WMNs needs to support more and more applications, such as voice over internet protocol (VoIP) and multimedia content distribution. To efficiently support these diverse demands and to effectively utilize the wireless network, service-oriented network layer design has been proposed recently. A key idea of such a service-oriented design is that a user only needs to specify the service to the network, instead of the destination address in a traditional manner. In this work, we present a simple solution, namely, the service-oriented routing algorithm (SORA), to enable service orientation in WMNs. The proposed scheme works in a similar manner as the Domain Name System (DNS). Specifically, SORA associates IP addresses to services just like the DNS associates IP addresses to domain names. To demonstrate the viability of the proposed scheme, we have developed simple but efficient programs that are running in Commercial off-the-shelf (COTS) devices, in particular, workstations and wireless mesh routers. Our experiments show that the proposed algorithm can efficiently provide distributed services in wireless mesh networks.

An Adaptive Cognition System for Smart Grids With Context Awareness and Fault Tolerance

In recent years, smart grids have been a common interest for many consumers, because of their comfort, safety, robustness, and economic characteristics. This paper presents the development of a computational tool, as an adaptive cognition system for smart grids, having smart homes as their composing nodes. Such a tool has been named Smart Home Energy Aware-Preserver (SHEAP). SHEAP incorporates evolutionary computation algorithms, and communication protocols, to provide users with context awareness and fault tolerance. Moreover, SHEAP considers a smart home powered by solar and wind energy, as a small version of the smart grid. SHEAP demonstrates the benefits of having a smart home that can control the amount of power needed, according to the context of usage. Furthermore, SHEAP includes fault tolerant mechanisms to monitor and react on fault occurrences. Simulation shows that with a smart control of the load, the requirements for a green energy system are reduced. An economic analysis of the approach demonstrates the viability of the project reducing the usage of grid energy by utilizing green energy.

A cognitive approach to load balancing for Green Houses

In this paper the case study of a smart home powered by solar and wind energy is presented. The benefits of having a smart home that can control the amount of power needed, according to the context of the usage, are also shown. Simulation shows that with a good control of the load it might be possible to reduce the installation costs of the Green Energy System. Furthermore, to support the results, a load balancing algorithm is created based on the Knapsack problem. An economic analysis of the approach is also shown to demonstrate the viability of the project, and how can the intelligence in the home lower the cost.

What defines an intruder? An intelligent approach

All attacks in a computer network begin with an intruders action of affecting the services provided to legitimate users. Hence, intrusion detection is vital for preserving integrity, confidentiality, and availability in a computer network. Intrusion detection faces many challenges, such as the need for large amount of data to discriminate between intruders and non-intruders, and the overlapping of user behavior to that of the intruders. This paper aims to target both of these challenges, by employing a distributed intrusion prevention system based on the Binary Partitle Swarm Optimization (BPSO) and Probabilistic Neural Network (PNN) algorithms. Such a system is capable of: 1) locally classifying actions as intruder or non-intruder type, and 2) consulting neighbors for casting a majority vote, upon finding high ambiguity on a decision. The algorithm uses an evolutionary computation approach to select the best features that can help classify intruders, while using fewer amounts of data. Furthermore, the approach uses concepts from semi-supervised learning to improve and adapt over time, to any network infrastructure. To demonstrate the viability of the proposed approach, a random set of data has been selected from the KDD-99 dataset. Such a set contained capture data from both users and attackers. Results have been compared with traditional data mining algorithms from previous work, demonstrating that such a system can have high accuracy, while maintaining a low false alarm rate.

Multimodal species identification in wireless sensor networks

This paper deals with a multimodal approach to identifying species in a Versatile Service-Oriented Wireless Mesh Sensor Network. This type of network is distinguished by the presence of heterogeneous networks, which may posses low storage capabilities. Hence, an optimal multimodal classifier is introduced, which employs audio and image features to enhance its performance in noisy environments. The classifier is a neural network which is evolved with an evolutionary algorithm. Results demonstrate that the classifier can achieve high performance, which is not degraded as it scales to classifying more classes.

A secured distribution of server loads to clients

Over the years, web services have become a fundamental part of our daily lives. We do everything on the web, from social to work and personal affairs. The volume of traffic over the web has become large due to the amount of people who use it. Recently, a paradigm for web applications called Single Page Application, has become more popular. With the use of Javascript based technologies, we can run practically any type of application on the client side, and leave the server with the sole duty of serving pages and a manual of the available content and how to use it, allowing the server to handle larger amount of volume. Unfortunately, SPA are also vulnerable to malicious users as most of its functionality is located at the client side. This paper proposed a framework to enhance the existent OAuth support to create a trust and anonymous service, which can help ensure the integrity of the application and the legitimacy of the user. Results show that SPA can not just be secure, but also help with server performance, by decreasing its load at higher traffic.

Gabor filter polynomial approximation based on a novel evolutionary stochastic technique

A new particle swarm optimization (PSO) algorithm has been developed, and combined with the differential evolution (DE) method. The novel evolutionary technique is utilized to approximate the sine and Gaussian functions of a Gabor filter, as polynomial functions, by the stochastic computation of an optimal set of coefficients. The new stochastic algorithm achieves a lower root mean square error of 0.0185, in comparison to sine and Gaussian approximations using state-machines from another work. Another important feature that adds more value to this work is the fact that polynomial functions can be constructed in hardware, through relatively simply operations, such as shift-add operations.

Adaptive strategies for Evolutionary Algorithm monitoring

Parameter tuning in Evolutionary Algorithms (EA), is a great obstacle that can become the key to success. Good parameter settings can yield optimal solutions, while bad settings may trap the EA, thus removing the chances of finding the optimal solutions. Therefore, it is vital that an optimal set of parameters configuration is chosen. It is a common practice to have a human expert that analyzes such parameters and modifies them accordingly. Such process is inefficient and expensive, since it requires time and is subject to human fatigue; it even becomes impractical if the environment is dynamic. This work proposes 2 adaptive strategies to tune such parameters: One Step Variation and a Fuzzy Logic Controller. A ranking scheme and modeling is introduced to evaluate the adaptive strategies. Results show that it may be possible to tune the parameters in an EA for achieving better results, without the need of an expert.