International Journal Information and Communication Technology Research
Volume:11 Issue: 4, Autumn 2019

Wireless sensor networks consist of a collection of sensor nodes deployed densely and randomly to fully cover a set of targets. Due to high redundancy incurred, it is possible to both preserve energy and enhance coverage quality by first switching off some sensors and then adjusting the sensing radius of the remaining ones. In this paper, the problem of target coverage in wireless sensor networks is studied by keeping a small number of active sensor nodes and adjusting the sensing radius of nodes. We propose a new game theory-based algorithm to target coverage. Inspired by current challenges in energy-limited sensor networks, we formulate the target coverage problem with adjustable sensing range as a repeated multiplayer game in which a utility function is formulated to consider the tradeoff between energy consumption and coverage quality. To solve the formulated game and achieve the Nash equilibrium, we present a distributed payoff based learning algorithm where each sensor only remembers its utility values and actions played during the last plays. The simulation results demonstrate the performance of our proposed game-theoretic algorithm and its superiority over previous approaches in terms of increasing the coverage rate and reducing the number of active nodes.

Given the complexity of today's networks, performing data analysis requires reducing the network’s size into smaller manageable useful sizes. To the best of our knowledge, in the domain of multilayer networks, reducing the size of such networks while simultaneously preserving the features and the nature of the network has not been done before. This paper, for the first time, combines three separate single-layer network simplification methods to make a new method for reducing the size of multilayer networks in a way that preserves the fundamental features of the network. The three simplification algorithms are Path Simplification, Degree-based Node Selection, and Hair Reduction algorithms. A hybrid approach is used for combining these algorithms with modifications to support multilayer features. To reduce the multilayer network, these algorithms are applied to the network sequentially. Our proposed method is tested on four real-world datasets. Results of the comparison among the reduced and the original networks, show that the reduced networks maintain the main features while their analysis complexity is less than the original ones.

In this paper, a secure chord protocol based method is presented to improve the latency and system storage requirements in smart metering. In the proposed approach, a secure multi-mode computation method is utilized which can reduce the time of data exchange and memory consumption, by maintaining data security and subscriber's privacy. This method can be utilized in smart metering networks based on the internet of things (IoT). According to the simulated results, the proposed method incremented the amount of production capacity by 26% compared to the reference model. Also, the average time to complete the data collection reduced by 65.5%, and the package delivery ratio of the proposed model incremented by 14.4 % in comparison with the reference model. Also, a secure mechanism-based lightweight authentication was provided. This scheme needs half memory usage versus other security plans such as the EDAS algorithms.

Nowadays, using Multi-Agent Systems (MASs) as a way of controlling complex and large-scale systems is becoming very popular. Also, since the scale of the systems is growing up and their environmental changes are becoming very fast and complicated, the experts are trying to enable these systems to control themselves, instead of having them controlled by humans. One way to devolve this responsibility to the systems is to use self-organization in MASs. To have a self-organizing MAS, agents should be able to shape up an organization. There are different organizational paradigms to be used in self-organizing MASs. Normally, the selection of organizational paradigm is done on design time by the designer of the system. But, in a rapidly changing and complicated environment, the selected paradigm might no longer be suitable for the system. In such a situation, there should be another way for the system to select a new suitable organizational paradigm at runtime. There are some works that provide a comparison among different organizational paradigms based on the performance of the MAS that uses that paradigm. But the comparison is done after the system is designed to have the paradigm. These works do not provide a mechanism for the system to select its paradigm at runtime. In this paper we propose an organization model for self-organizing MASs that provides these systems with the ability to change their organizational paradigm at run time. This organization model considers an amount of utility as the criterion based on which the currently used paradigm can be changed by the system itself. We simulate and evaluate our model in an IoT scenario. The scenario includes a Smart Home and its application of Comfort of Residences. The results show that changing the organizational paradigm and not sticking to the selected one on design time, gives us a 28% improvement on the utility.

The popularity of social networks has rapidly increased over the past few years. Social networks provide many kinds of services and benefits to their users like helping them to communicate, click, view and share contents that reflect their opinions or interests. Detecting important contents defined as the most visited posts and users whom disseminate them can provide some interesting insights from cyberspace user’s activities. In this paper, a framework for discovering important posts (most popular posts by views count) and influential users is introduced. The proposed framework employed on Telegram instant messaging service in this study but it is also applicable to other social networks such as Instagram and Twitter. This framework continuously works in a real social network analysis system named Zekavat to find daily important posts and influential users. The effectiveness of this framework was shown in experiments. The accuracy achieved in the advertisement detection model is 89%. Text-based clustering part of the framework was tested based on the human factor verification and clustering time is less than linear. Graph creation based on publishing relationships is more effective than mention relationship and in this process influential users can be identified in a precise manner.

Community detection is one of the important topics in complex network study. There are many algorithms for community detection, some of which are based on finding maximal cliques in the network. In this paper, we improve Streaming Community Detection Algorithm (SCoDA) and Order Statistics Local Optimization Method (OSLOM). After finding maximal cliques and generating the corresponding graphs, the latter are used as input to SCoDA and OSLOM algorithms. Non-overlap and overlap synthetic graphs and real graphs data are used in our experiments.  As evaluation criteria F1score and NMI scores functions are utilized. It is shown that the improved version of SCoDA has better results in comparison to the original SCoDA algorithm, and the improved OSLOM algorithm has better performance in comparison with the original OSLOM algorithm.

Text classification has a wide range of applications such as: spam filtering, automated indexing of scientific articles, identification the genre of documents, news monitoring, and so on.  Text datasets usually contain much irrelevant and noisy information which eventually reduces the efficiency and cost of their classification. Therefore, for effective text classification, feature selection methods are widely used to handle the high dimensionality of data. In this paper, a novel feature selection method based on the combination of information gain and FAST algorithm is proposed. In our proposed method, at first, the information gain is calculated for the features and those with higher information gain are selected. The FAST algorithm is then used on the selected features which uses graph-theoretic clustering methods. To evaluate the performance of the proposed method, we carry out experiments on three text datasets and compare our algorithm with several feature selection techniques. The results confirm that the proposed method produces smaller feature subset in shorter time. I addition, The evaluation of a K-nearest neighborhood classifier on validation data show that, the novel algorithm gives higher classification accuracy.