Application of neural networks and machine learning in networkdesign

Communication network design is becoming increasingly complex, involving making networks more usable, affordable, and reliable. To help with this, we have proposed an expert network designer (END) for configuring, modeling, simulating, and evaluating large structured computer networks, employing artificial intelligence, knowledge representation, and network simulation tools. We present a neural network/knowledge acquisition machine-learning approach to improve the END's efficiency in solving the network design problem and to extend its scope to acquire new networking technologies, learn new network design techniques, and update the specifications of existing technologies     

Multifold node authentication in mobile ad hoc networks

An ad hoc network is a collection of nodes that do not need to rely on a predefined infrastructure to keep the network connected. Nodes communicate amongst each other using wireless radios and operate by following a peer‐to‐peer network model. In this article, we propose a multifold node authentication approach for protecting mobile ad hoc networks. The security requirements for protecting data link and network layers are identified and the design criteria for creating secure ad hoc networks using multiple authentication protocols are analysed. Such protocols, which are based on zero‐knowledge and challenge‐response techniques, are presented through proofs and simulation results. Copyright © 2007 John Wiley & Sons, Ltd.     

Fuzzy generalized network approach for solving an optimization model for routing in B‐ISDN

Routing algorithms are required to guarantee the various quality of service (QoS) characteristics requested by the wide range of applications supported by Broadband Integrated Services Digital Networks (B‐ISDN). In this paper the routing problem is formulated as a fuzzy multiobjective optimization model. The fuzzy approach allows for the inclusion and evaluation of several criteria simultaneously. The proposed model takes into consideration the balancing of the load in the network to avoid link saturation and hence the possibility of congestion. A hybrid approach that combines the generalized network concept with the technique of fuzzy programming is recommended to solve the model. The efficiency and applicability of the model is tested under different load conditions by studying several measures of performance. This revised version was published online in June 2006 with corrections to the Cover Date.     

Data link layer design issues in a home automation system

In this paper, we examine design issues relevant to the implementation of a data link layer of a home automation system based on the Electronics Industries Association (EIA) Consumer Electronics Bus (CEBus). The hardware and software components of the system are described along with several techniques that enhance the performance of the system, allow for a faster implementation using off-the-shelf components and provide low overhead. The proposed design has been incorporated into a basic home automation system (BHAS) that is used as a testbed for alternative technologies and algorithms.     

Improved estimator for a discretised learning routing algorithm

The use of learning automata in dynamic routing algorithms has shown promise. An improved estimator for the new scheme flow control with discretised learning (NSDL) routing algorithm is presented which provides a better profile of the network environment, to improve the performance of the NSDL scheme.<>     

LIDF: Layered intrusion detection framework for ad-hoc networks

As ad-hoc networks have different characteristics from a wired network, the intrusion detection techniques used for wired networks are no longer sufficient and effective when adapted directly to a wireless ad-hoc network. In this article, first the security challenges in intrusion detection for ad-hoc networks are identified and the related work for anomaly detection is discussed. We then propose a layered intrusion detection framework, which consists of collection, detection and alert modules that are handled by local agents. The collection, detection and alert modules are uniquely enabled with the main operations of ad-hoc networking, which are found at the OSI link and network layers. The proposed modules are based on interpolating polynomials and linear threshold schemes. An experimental evaluation of these modules shows their efficiency for several attack scenarios, such as route logic compromise, traffic patterns distortion and denial of service attacks.     

An integrated framework for managing emergency-response logistics:the case of the electric utility companies

Managers of emergency response operations in electric utility companies frequently face decisions regarding the optimum deployment of the available emergency response resources. The complexity of emergency repair problems requires the development of a decision-making framework able to address strategic and tactical emergency repair issues. This paper presents an integrated framework consisting of a data management module, a vehicle monitoring and communications module and a modeling module. The functional characteristics of the proposed system are analyzed in the context of emergency repair operations of an electric utility company. Also presented is an extensive case study encompassing the three modules of the proposed decision-making system and showing the improvement in effectiveness of emergency response operations by introducing operations research techniques coupled with advanced information technologies     

A telecommunications quality study using the analytic hierarchyprocess

An analytic hierarchy process (AHP) approach that can be used either for the analysis and comparison of the quality of several telecommunications companies, or for the evaluation of alternate technologies in telecommunications is presented. The method is based on pairwise comparison between several factors that affect the quality of service in a hierarchical structure. Two particular formulations are presented and discussed extensively     

A location-based mobility tracking scheme for PCS networks

This paper introduces a location-based locating strategy for Personal Communication Services (PCS) systems. In the proposed scheme, location updates are based on the value of a movement counter. The update of the movement counter is based on information on the locations visited by a mobile terminal (MT) since the last location update. In particular, the cell identifiers and the corresponding movement counter values are maintained in the MT for a part of the visited cells since the last location update. When the MT enters a cell whose identifier is not stored in the MT, the movement counter increases by one. If the movement counter reaches the update threshold, a location update is triggered; otherwise, the cell identifier and the corresponding movement counter value are stored in the MT. When the MT enters a cell whose identifier is in the MT, the movement counter is assigned with the counter value kept in the MT for this cell, and all cells with greater counter values in the MT are removed. The proposed scheme allows the dynamic selection of the update threshold according to each user's calling and mobility patterns. Analytical and simulation models have been developed to compare the proposed scheme with both the movement-based scheme and the distance-based scheme. Results demonstrate that when the call-to-mobility ratio (CMR) is low, the proposed scheme can achieve a significant cost reduction compared to the movement-based scheme. For example, when the CMR is 0.01, a savings of around 25% in the optimal total cost per call arrival is achieved with the proposed scheme. Moreover, the proposed scheme is easier to implement than the distance-based scheme.