Ahmed Sameh

Mobile user movement prediction using bayesian learning for neural networks

A technique for reducing the wireless cost of tracking mobile users with uncertain parameters is developed in this paper. Such uncertainty arises naturally in wireless networks, since an efficient user tracking is based on a prediction of its future call and mobility parameters. The conventional approach based on dynamic tracking is not reliable in the sense that inaccurate prediction of the user mobility parameters may significantly reduce the tracking efficiency. Unfortunately, such uncertainty is unavoidable for mobile users, especially for a burst mobility patterns. In this paper, we present a novel hybrid Bayesian neural network model for predicting locations on Cellular Networks (can also be extended to other wireless networks such as WI-FI and WiMAX). We investigate different parallel implementation techniques on mobile devices of the proposed approach and compare it to many standard neural network techniques such as: Back-propagation, Elman, Resilient, Levenberg-Marqudat, and One-Step Secant models. Bayesian learning for Neural Networks predicts location better than standard neural network techniques since it uses well founded probability model to represent uncertainty about the relationships being learned. The result of Bayesian training is a posterior distribution over network weights. We use Markov chain Monte Carlo methods (MCMC) to sample N values from the posterior weights distribution.

Movement Prediction Using Bayesian Learning for Neural Networks

A technique for reducing the wireless cost of tracking mobile users with uncertain parameters is developed in this paper. Such uncertainty arises naturally in wireless networks, since an efficient user tracking is based on a prediction of its future call and mobility parameters. The conventional approach based on dynamic tracking is not reliable in the sense that inaccurate prediction of the user mobility parameters may significantly reduce the tracking efficiency. Unfortunately, such uncertainty is unavoidable for mobile users, especially for a burst mobility patterns. In this paper, we present a novel hybrid Bayesian neural network model for predicting locations on Cellular Networks (can also be extended to other wireless networks such as WI-FI and WiMAX). We investigate different parallel implementation techniques on mobile devices of the proposed approach and compare it to many standard neural network techniques such as: Back-propagation, Elman, Resilient, Levenberg-Marqudat, and One-Step Secant models. Bayesian learning for Neural Networks predicts location better than standard neural network techniques since it uses well founded probability model to represent uncertainty about the relationships being learned. The result of Bayesian training is a posterior distribution over network weights. We use Markov chain Monte Carlo methods (MCMC) to sample N values from the posterior weights distribution.

Bayesian Learning of Neural Networks for Mobile User Position Prediction

In this paper, a novel technique for location prediction of mobile users has been proposed, and a paging technique based on it is developed. Mobile users are creatures of habits. They tend to repeat their behaviors. Hence, neural networks with its learning and generalization ability may act as a suitable tool to predict the location of a mobile user provided it is trained appropriately by the personal mobility profile. For prediction, a novel hybrid Bayesian neural network model for predicting locations on Cellular Networks (can also be extended to other wireless networks such as Wi-Fi and WiMAX) is suggested. We investigate its different parallel implementation techniques on mobile devices, and compare its performance to many standard neural network techniques such as: Back-propagation, Elman, Resilient, Levenberg-Marqudat, and One-Step Secant models. This approach is free from all unrealistic assumptions about the movement of the users. It is applicable to any arbitrary cell architecture. It attempts to reduce the total location management cost and paging delay. In general, it enhances mobility management in wireless networks (in location management and hand-off management). In our experiments, we compare results of the proposed Bayesian Neural Network with 5 standard neural network techniques in predicting next location. Bayesian learning for Neural Networks predicts location better than standard neural network techniques since it uses well founded probability model to represent uncertainty about the relationship being learned. The result of Bayesian training is a posterior distribution over network weights.

Modeling Jini-UPnP Bridge using Rapide ADL

The exploding deployment of network enabled mobile devices, along with the expansion of networked services have created the need for users to easily manage these devices and services and also to coordinate with one another. Service Discovery Protocol (SDP) enables networked devices, applications, and services to seek out and find other complementary networked devices, applications, and services needed to properly complete specified tasks. A number of bridging techniques have been proposed and implemented. In this paper we propose a one directional bridging system (Jini-UPnP Bridge). To validate the proposed system, we model and simulate the bridge using Rapide ADL simulation and analysis toolset. We perform a number of simulation tests and use the Rapide Poset viewer to analyze the simulators output Poset tree of events.

Dealing with Quality of Service in Hybrid Wired-Wireless Networks

The Quality of Service (QoS) research investigations in Wired and Wireless networks have been conducted mostly in isolation [2]. Recently, a need for an end to end quality of service over hybrid networks (containing wired and wireless segments) has become evident. IEEE 802.11e work group has set up the standards for wireless network quality of service which became part of the comprehensive approved 802.11 wireless networks standards late 2006. The most recent version of IEEE 802.11 standard states that there is plenty of room for improvement and development in the area of integrating QoS in wireless and wired networks. On the other hand, Resource Reservation Protocol (RSVP) is widely accepted to be the complex of all QoS technologies on wired networks. It ensures, to a great extent, the quality of any given connection on the network. Some researches have been done to extend RSVP to wireless networks. Moreover, RSVP-TE is an extension of RSVP that allows the establishing of MPLS (Multi Protocol Label Switching) LSPs (Label Switching Paths) in the network. MPLS is a newer technology that offers service integration, layer 2 switching and connection orientedness, that allows traffic engineering controlled flows in the network. MPLS is a technology that delivers a unified control mechanism with its multiprotocol capabilities for running over mixed media infrastructures. MPLS defines signaling mechanisms to support both Class of Service (CoS) and QoS. In this paper we are investigating how the MPLS hierarchical architecture for label-switched networks can be used for supporting wireless users. This architecture involves requirements at the mobile terminal for initiating or hopping label switched paths at the air interface, and allowing end to end interconnection to the backbone network. We are utilizing aspects of the 802.11 QoS standards and techniques proposed in recent research to propose a technique to extend RSVP-TE into the WLANs (aka Wi-Fi) domain using MPLS.

Security in mobile agent systems

There are four possible security attacks: Attacks from host to host, from agents to hosts, from agents to agents and finally from hosts to agents. Our main concern is attacks from a malicious host on an agent. These attacks can take many forms including rerouting, spying out code, spying out data, spying out control flow, manipulation of code, manipulation of data, manipulation of control flow, incorrect execution of code, masquerading and denial of execution. This paper presents a three-tier solution of code mess up, encryption and time out.

System-Level Modeling and Design Using SysML and SystemC

SysML, the dedicated system-level UML-based notations proposed by the OMG, is gaining a lot of momentum to be widely accepted by system-level designers. IEEE has standardized SystemC to gain the fame of being powerful at the system-level architectural design and verification. Building on both SystemC and SysML success, we have dedicated this research to study and prototype the automatic translation of SysML designs into SystemC models. This work is an attempt to accelerate the SoC design process by raising the abstraction level in an automated environment.

Ontology-Based Feedback E-Learning System for Mobile Computing

An E-Learning system that provides vast quantities of annotated resources (fragments or learning objects) and produces semantically rich feedback is very desirable. It is an accepted psychological principle that some of the essential elements needed for effective learning are custom learning and semantic feedback. In this paper we are making use of a collection (ontology) of metadata for the design of a custom E-Learning system that also provides learners with effective semantic educational feedback support. The learning domain is “mobile computing.” We define various concepts in the domain and the relationships among them as the ontology, and build a system to utilize them in customizing the E-Learning process. The ontology is also used to provide informative feedback from the system during learning and/or during assessment. The focus in this research is on the representation of ontology using languages/grammars, grammar analysis techniques, algorithms, and AI mechanisms to customize the learning and create effective feedbacks. The proposed mechanisms, based on ontology, are used to assemble virtual courses and create a rich supply of feedbacks, not only in assessment situations but also in the context of design-oriented education. We are targeting feedbacks similar to ones in programming environments and design editors.

The Context Oriented Architecture: Integrating Context into Semantic Web Services

While emerging standards promote the interoperability of Web services, Web services adaptability to user and application context is a requirement yet to be investigated. Web services today encapsulate application and business logic and are candidates for supporting pervasive environments, where adaptability will have added value in all of these domains. In this paper we introduce the context oriented architecture; an architecture for building context aware applications using Web services as building components. The context oriented architecture uses semantic description to automate monitoring and responding to service context. We extend the specifications and implementation of the semantic Web service ontology language (OWL-S) to represent the context profile of a Web service

The Context Oriented Architecture: An Augmentation of Context Awareness and Reactivity into Web Services

Standardization promotes web services as a very promising candidate for successfully integrating disparately heterogeneous systems. As such, web services prove themselves to be very suitable candidates for supporting the predominantly heterogeneous pervasive environments. The ability of web services however to sense their surrounding context and effectively react to it is still a matter of research. In this work, we introduce a new architecture, an architecture solely built on open standards, that supports the development of context aware and context reactive applications that use web services as building components. We describe in detail the various components of this architecture, along with their supporting interactions. Furthermore, we describe the expansion of the OWL-S ontology language, namely expanding both the profile and the process model ontologies, to allow for encoding context behavior of both web services and clients. We eventually illustrate validation scenarios for this architecture, and demonstrate an application example built using this architecture that adapts to ambient security requirements.