Recent advances in radio environmental mapping enable novel, practical and efficient cognitive radio and dynamic spectrum access solutions. A crucial aspect of such solutions is to ensure the reliability of the constructed Radio Environmental Maps (REMs). Especially important is the accurate and up-todate Radio Interference Field (RIF) estimation based on distributed spectrum use measurements. This paper analyzes the use of spatial interpolation techniques that allow robust, yet sufficiently reliable, RIF estimation from a limited number of field measurements. Several spatial interpolation techniques based on Inverse Distance Weighting (IDW) are analyzed and compared in terms of reliability bounds of the interpolation errors for an indoor environment. Performance evaluation using REM prototype implementation and a wireless testbed shows that the spatial interpolation techniques can provide a robust and reliable RIF estimation within the entire REM concept. Keywords-Radio Environmental Map (REM), Radio Interference Field (RIF), spatial interpolation, IDW, reliability.I. II. REM DESIGN AND PROTOTYPE IMPLEMENTATIONREMs are a fundamental enabling technology to implement practical cognitive radio networks and dynamic spectrum access solutions. This section presents a general REM architecture and prototype, the REM data model and representation with special focus on the radio interference field (RIF) estimation through spatial interpolation. A. REM ArchitectureA general architecture for REM generation and evaluation should comprise several key functionalities, starting from the spectrum measurements execution and data acquisition, through the data processing and REM construction, to the REM data presentation and utilization for various spectrum WK ,QWHUQDWLRQDO ,&67 &RQIHUHQFH RQ &RJQLWLYH 5DGLR 2ULHQWHG :LUHOHVV 1HWZRUNV DQG &RPPXQLFDWLRQV &52:1&20 k ,&67 '2, LFVWFURZQFRP management purposes. In order to satisfy these requirements, a functional REM architecture (Fig. 1) should consist of (at least) the following functional entities [6]: Measurement Capable Devices, REM data Storage and Acquisition unit, REM Manager and a REM User. The subsequent paragraphs briefly explain these entities and the respective interfaces. REM Storage and Acquisition unit (REM SA)

Abstract-Cognitive radio networks challenge the traditional wireless networking paradigm by introducing concepts firmly stemmed into the Artificial Intelligence (AI) field, i.e., learning and reasoning. This fosters optimal resource usage and management allowing a plethora of potential applications such as secondary spectrum access, cognitive wireless backbones, cognitive machine-to-machine etc. The majority of overview works in the field of cognitive radio networks deal with the notions of observation and adaptations, which are not a distinguished cognitive radio networking aspect. Therefore, this paper provides insight into the mechanisms for obtaining and inferring knowledge that clearly set apart the cognitive radio networks from other wireless solutions.

No abstract

The paper presents a new method for loss allocation in radial distribution networks (DN) with dispersed generation (DG). The method is based on the power summation algorithm, the branch-oriented approach in particular, and establishes direct relationship between losses in each branch of the network and injected active and reactive power in the nodes. Active and reactive power at receiving node of each branch of the network is decomposed as a sum of all node injected powers, including branch losses in the branches fed by the particular branch. Allocation of crossed terms of active and reactive power is calculated using quadratic relation between power flows and losses. DG active and reactive power injections in the summation process are treated as negative loads, while active and reactive loads are considered positive. Results of the application of the method on 32-node test system are presented, discussed, and compared with other published methods.