Vol. 32 · No. 6

The aim of the research was to develop an Unmanned Aerial Vehicle (UAV) hardware and software which can control a small size UAV (with 1–10 kg takeoff weight) from takeoff to landing. The research was based on AERObot V5 autopilot, developed and tested previously by us. AERObot can control a chosen aerial vehicle on a predefined route and altitude reliably but requires manual takeoff and landing. The speciality of the new takeoff and landing methods is that they do not need any additional RF guidance transmitter, onboard radio/ultrasonic altimeter or special navigation marker points. Using only the existing onboard sensors (GPS, Inertial Measurement Unit – IMU, barometric airspeed meter, barometric altimeter) AERObot can safely take off and land the UAV under the specified weather limitations for the aerial vehicle frame.

A new approach of the string theory called the Prediction Model Based on String Invariants (PMBSI) was applied here to time-series forecast. We used 2-end-point open string that satisfies the Dirichlet and Neumann boundary conditions. The initial motivation was to transfer modern physical ideas into the neighboring field called econophysics. The physical statistical viewpoint has proved to be fruitful, namely in the description of systems where many-body effects dominate. However, PMBSI is not limited to financial forecast. The main advantage of PMBSI includes absence of the learning phase when large number of parameters must be set. Comparative experimental analysis of PMBSI vs. SVM was performed and the results on artificial and real-world data are presented. PMBSI performance was in a close match with SVM.

This article introduces a statistical approach for the maximization of the financial benefits yielded by software projects that have been broken down into a considerable number of minimum marketable features modules (MMFs) and architectural elements (AEs). As the statistical approach requires a polynomial computational effort to run and provides approximation solutions with an arbitrarily chosen degree of confidence, it allows managers and developers to be more confident about the rightness of the decisions they make with little additional computational effort.

Clustering spatial data is a well-known problem that has been extensively studied. In the real world, there are many physical obstacles such as rivers, lakes, highways, and mountains, whose presence may substantially affect the clustering result. Although many methods have been proposed in previous works, very few have considered physical obstacles and interlinking bridges. Taking these constraints into account during the clustering process is costly, yet modeling the constraints is paramount for good performance. Owing to saturation in existing telephone networks and the ever increasing demand for wire and wireless services, telecommunication engineers are looking at technologies that can deliver sites and satisfy the demand and level of service constraints in an area with and without obstacles. In this paper, we study the problem of clustering in the presence of obstacles to solve the network planning problem. As such, we modified the NetPlan algorithm and developed the COD-NETPLAN (Clustering with Obstructed Distance – Network Planning) algorithm to solve the problem of 2D and 3D obstacles. We studied the problem of determining the location of the multi service access node in an area with many mountains and rivers. We used a reachability matrix to detect 2D obstacles, and line segment intersection together with geographical information system techniques for 3D obstacles. Experimental results and the subsequent analysis indicate that the COD-NETPLAN algorithm is both efficient and effective.

Recently, information leakage incidents have occurred due to database security vulnerabilities. The administrators in the traditional database access control methods grant simple permissions to users for accessing database objects. Even though they tried to apply more strict permissions in recent database systems, it was difficult to properly adopt sophisticated access control policies to commercial databases due to performance degradations. This paper proposes a database security system including a database firewall server as an enhanced database access control system which can efficiently enforce sophisticated security policies to provide database with confidentiality using a data masking technique for diverse conditions such as the date, time, SQL string, and table columns to database systems.

This paper proposes a method of automatic facial reconstruction from a facial image partially corrupted by noise or occlusion. There are two key features of this method; the one is the automatic extraction of the correspondences between the corrupted input face and reference face without additional manual tasks; the other is the reconstruction of the complete facial information from corrupted facial information based on these correspondences. In this paper, we propose a non-iterative approach that can match multiple feature points in order to obtain the correspondences between the input image and the reference face. Furthermore, shape and texture of the whole face are reconstructed by SVDD (Support Vector Data Description) from the partial correspondences obtained by matching. The experimental results of facial image reconstructions show that the proposed SVDD-based reconstruction method gives smaller reconstruction errors for a facial image corrupted by Gaussian noise and occlusion than the existing linear projection reconstruction method with a regulation factor. The proposed method also reduces the mean intensity error per pixel by an average of 35

The current Internet is facing serious scalability problems and the overloading of Internet Protocol (IP) addresses is regarded as an important reason. The Locator/ID Separation Protocol (LISP) is proposed as a network-based solution that separates IP addresses into Routing Locators (RLOCs) and Endpoint Identifiers (EIDs) to address the routing scalability problems. It is a critical challenge for LISP to design a scalable and efficient mapping system. In this paper, we propose a hierarchical mapping system (HMS). HMS consists of two levels with the bottom level maintaining the EID-to-RLOC mappings in an Autonomous System (AS) and the upper level storing the mappings between EID-prefixes and ASs in the global network. We adopt one-hop Distributed Hash Table (DHT) to organize EID-to-RLOC mappings in the bottom level and use a protocol like Border Gateway Protocol (BGP) to propagate EID-prefix-to-AS mappings in the upper level. HMS aggregates the prefixes in an AS and decreases the global mapping entries in the upper level. The evaluation results show that the number of mapping entries in HMS grows slower than the routing table size, which makes HMS scalable. In addition, the mobility in HMS does not cause mapping changes in the upper level. It makes HMS efficient in supporting host mobility. We estimate the map-requests sent to the mapping system, which show the load on HMS is small. Last, we compare HMS with LISP-TREE and LISP+ALT by quantitative analysis, in terms of resolution cost, and qualitative analysis. The results show that HMS has a good performance.

A preliminary combination of Standard Model Higgs searches with the ATLAS experiment, in a dataset collected at energy sqrts = 7 TeV at the LHC in year 2011, is presented. footnoteThe paper refers to the status of Higgs boson searches as from March 2012, before the Higgs discovery at LHC. The Higgs boson mass ranges from 112.7 GeV to 115.5 GeV, 131 GeV to 237 GeV and 251 GeV to 468 GeV are excluded at the 95

Monte Carlo simulations are widely used e.g. in the field of physics and molecular modelling. The main role played in these is by the high performance random number generators, such as RANLUX or MERSSENE TWISTER. In this paper the authors introduce the world's first implementation of the RANLUX algorithm on an FPGA platform for high performance computing purposes. A significant speed-up of one generator instance over 60 times, compared with a graphic card based solution, can be noticed. Comparisons with concurrent solutions were made and are also presented. The proposed solution has an extremely low power demand, consuming less than 2.5 Watts per RANLUX core, which makes it perfect for use in environment friendly and energy-efficient supercomputing solutions and embedded systems.

The tracking and synchronization problem of uncertain chaotic system, which is considered to be applied in secure communication in the future by many researchers, is considered in this paper. A double integral sliding mode controller is adopted to cope with the uncertainties of the chaotic system. Adaptive and robust strategies, such as Nussbaum gain method, are used to solve the unmodeled dynamic problem and unknown control direction problem. Meanwhile, the stability of the whole system is guaranteed by constructing of a big Lyapunov function for the whole system. Finally, a four dimension super-chaotic system is used as an example to do the numerical simulation and it testifies the rightness and effectiveness of the proposed method.

Today's most common firewalls are mostly rule-based. Their knowledge consists of a set of rules upon which they process received packets. They cannot do anything they have not been explicitly configured to do. This makes the system more straightforward to set up, but less flexible and less adaptive to changing circumstances. We will investigate a network firewall whose rule-base we will try to model using an artificial neural network, more specifically using a multi-layer perceptron (MLP) trained by the back-propagation algorithm. The training data are acquired from the network and we consider two possible scenarios. In Scenario 1, the user has no firewall available and the policy is deduced from the existing traffic in the network which is considered to be legitimate. In Scenario 2, the learning module is placed behind the existing firewall (or firewalls) in order to learn their behavior. In both cases, all traffic, which is recorded, contains only positive examples; however, a direct training of a MLP from a set of positive examples is impossible. We solved this problem using a synthetic generation of negative examples which led to successful learning.