Ahmed Ghoneim

Software defined healthcare networks

With the increasingly serious problem of the aging population, creating an efficient and real-time health management and feedback system based on the healthcare Internet of Things (HealthIoT) is an urgent need. Specifically, wearable technology and robotics can enable a user to collect the required human signals in a comfortable way. HealthIoT is the basic infrastructure for realizing health surveillance, and should be flexible to support multiple application demands and facilitate the management of infrastructure. Therefore, enlightened by the software defined network, we put forward a smart healthcare oriented control method to software define health monitoring in order to make the network more elastic. In this article, we design a centralized controller to manage physical devices and provide an interface for data collection, transmission, and processing to develop a more flexible health surveillance application that is full of personalization. With these distinguished characteristics, various applications can coexist in the shared infrastructure, and each application can demand that the controller customize its own data collection, transmission, and processing as required, and pass the specific configuration of the physical device. This article discusses the background, advantages, and design details of the architecture proposed, which is achieved by an open-ended question and a potential solution. It opens a new research direction of HealthIoT and smart homes.

Green Video Transmission in the Mobile Cloud Networks

Video transmission is an indispensable component of most applications related to the mobile cloud networks (MCNs). However, because of the complexity of the communication environment and the limitation of resources, attempts to develop an effective solution for video transmission in the MCN face certain difficulties. In this paper, we propose a novel green video transmission (GVT) algorithm that uses video clustering and channel assignment to assist in video transmission. A video clustering model is designed based on game theory to classify the different video parts stored in mobile devices. Using the results of video clustering, the GVT algorithm provides the function of channel assignment, and its assignment process depends on the content of the video to improve channel utilization in the MCN. Extensive simulations are carried out to evaluate the GVT with several performance criteria. Our analysis and simulations show that the proposed GTV demonstrates a superior video transmission performance compared with the existing methods.

Big Data-Driven Service Composition Using Parallel Clustered Particle Swarm Optimization in Mobile Environment

The proliferation of mobile computing and smartphone technologies has resulted in an increasing number and range of services from myriad service providers. These mobile service providers support numerous emerging services with differing quality metrics but similar functionality. Facilitating an automated service workflow requires fast selection and composition of services from the services pool. The mobile environment is ambient and dynamic in nature, requiring more efficient techniques to deliver the required service composition promptly to users. Selecting the optimum required services in a minimal time from the numerous sets of dynamic services is a challenge. This work addresses the challenge as an optimization problem. An algorithm is developed by combining particle swarm optimization and k-means clustering. It runs in parallel using MapReduce in the Hadoop platform. By using parallel processing, the optimum service composition is obtained in significantly less time than alternative algorithms. This is essential for handling large amounts of heterogeneous data and services from various sources in the mobile environment. The suitability of this proposed approach for big data-driven service composition is validated through modeling and simulation.

A Triaxial Accelerometer-Based Human Activity Recognition via EEMD-Based Features and Game-Theory-Based Feature Selection

In recent years, sensor-based human activity recognition has attracted lots of studies. This paper presents a single wearable triaxial accelerometer-based human activity recognition system, which can be used in the real life of activity monitoring. The sensor is attached around different parts of the body: waist and left ankle, respectively. In order to improve the accuracy and reduce the computational complexity, the ensemble empirical mode decomposition (EEMD)-based features and the feature selection (FS) method are introduced, respectively. Considering the feature interaction, a game theory-based FS method is proposed to evaluate the features. Relevant and distinguished features that are robust to the placement of sensors are selected. In the experiment, the data acquired from the two different parts of the body, waist and ankle, are utilized to evaluate the proposed FS method. To verify the effectiveness of the proposed method, k-nearst neighbor and support vector machine are used to recognize the human activities from waist and ankle. Experiment results demonstrate the effectiveness of the introduced EEMD-based features for human activity recognition. Compared with the representative FS methods, including Relief-F and minimum-redundancy maximum-relevance, the proposed FS approach selects fewer features and provides higher accuracy. The results also show that the triaxial accelerometer around the waist produces optimal results.

Folksonomy-Based Visual Ontology Construction and Its Applications

An ontology hierarchically encodes concepts and concept relationships, and has a variety of applications such as semantic understanding and information retrieval. Previous work for building ontologies has primarily relied on labor-intensive human contributions or focused on text-based extraction. In this paper, we consider the problem of automatically constructing a folksonomy-based visual ontology (FBVO) from the user-generated annotated images. A systematic framework is proposed consisting of three stages as concept discovery, concept relationship extraction, and concept hierarchy construction. The noisy issues of the user-generated tags are carefully addressed to guarantee the quality of derived FBVO. The constructed FBVO finally consists of 139 825 concept nodes and millions of concept relationships by mining more than 2.4 million Flickr images. Experimental evaluations show that the derived FBVO is of high quality and consistent with human perception. We further demonstrate the utility of the derived FBVO in applications of complex visual recognition and exploratory image search.

Localization Based on Social Big Data Analysis in the Vehicular Networks

Location-based services, especially for vehicular localization, are an indispensable component of most technologies and applications related to the vehicular networks. However, because of the randomness of the vehicle movement and the complexity of a driving environment, attempts to develop an effective localization solution face certain difficulties. In this paper, an overlapping and hierarchical social clustering model (OHSC) is first designed to classify the vehicles into different social clusters by exploring the social relationship between them. By using the results of the OHSC model, we propose a social-based localization algorithm (SBL) that use location prediction to assist in global localization in the vehicular networks. The experiment results validate the performance of the OHSC model and show that the presented SBL algorithm demonstrates superior localization performance compared with the existing methods.

Head Pose Estimation on Top of Haar-Like Face Detection: A Study Using the Kinect Sensor

Head pose estimation is a crucial initial task for human face analysis, which is employed in several computer vision systems, such as: facial expression recognition, head gesture recognition, yawn detection, etc. In this work, we propose a frame-based approach to estimate the head pose on top of the Viola and Jones (VJ) Haar-like face detector. Several appearance and depth-based feature types are employed for the pose estimation, where comparisons between them in terms of accuracy and speed are presented. It is clearly shown through this work that using the depth data, we improve the accuracy of the head pose estimation. Additionally, we can spot positive detections, faces in profile views detected by the frontal model, that are wrongly cropped due to background disturbances. We introduce a new depth-based feature descriptor that provides competitive estimation results with a lower computation time. Evaluation on a benchmark Kinect database shows that the histogram of oriented gradients and the developed depth-based features are more distinctive for the head pose estimation, where they compare favorably to the current state-of-the-art approaches. Using a concatenation of the aforementioned feature types, we achieved a head pose estimation with average errors not exceeding 5.1∘,4.6∘,4.2∘ for pitch, yaw and roll angles, respectively.

Deep Relative Attributes

Relative attribute (RA) learning aims to learn the ranking function describing the relative strength of the attribute. Most of current learning approaches learn a linear ranking function for each attribute by use of the hand-crafted visual features. Different from the existing study, in this paper, we propose a novel deep relative attributes (DRA) algorithm to learn visual features and the effective nonlinear ranking function to describe the RA of image pairs in a unified framework. Here, visual features and the ranking function are learned jointly, and they can benefit each other. The proposed DRA model is comprised of five convolutional neural layers, five fully connected layers, and a relative loss function which contains the contrastive constraint and the similar constraint corresponding to the ordered image pairs and the unordered image pairs, respectively. To train the DRA model effectively, we make use of the transferred knowledge from the large scale visual recognition on ImageNet [1] to the RA learning task. We evaluate the proposed DRA model on three widely used datasets. Extensive experimental results demonstrate that the proposed DRA model consistently and significantly outperforms the state-of-the-art RA learning methods. On the public OSR, PubFig, and Shoes datasets, compared with the previous RA learning results [2], the average ranking accuracies have been significantly improved by about 8%, 9%, and 14%, respectively.

From the Service-Oriented Architecture to the Web API Economy

As Web APIs become the backbone of Web, cloud, mobile, and machine learning applications, the services computing community will need to expand and embrace opportunities and challenges from these domains.

QoS management for dependable sensory environments

Sensory environments for healthcare are commonplace nowadays. A patient monitoring system in such an environment deals with sensor data capture, transmission and processing in order to provide on-the-spot support for monitoring the vulnerable and critical patients. A fault in such a system can be hazardous on the health of the patient. Therefore, such a system must be dependable and ensure reliability, fault-tolerance, safety and other critical aspects, in order to deploy it in real scenario. Also, the management of the infrastructure resources must be efficient and the eventual system reconfiguration must be reliably performed. This paper encounters some of these issues and proposes a component platform with specific support for several QoS aspects, namely fault tolerance, safe inter-component communication and resource management. The platform adopts the Service Component Architecture (SCA) model and defines a Data Distribution Service (DDS) binding, which provides the fault tolerance and the required safety-ensuring techniques and measures, as defined in the IEC 61784-3-3 standard. As a proof of concept, a distributed home care application that improves the medical assistance in case of fire detection is presented.