Abeer Hamdy

An Efficient Workload Allocation to Improve Scheduling Real-Time Tasks

Cluster computing systems are currently used to execute high performance applications that exhibit real time characteristics. A high-speed communication networks are used to interconnect the nodes of a cluster. Also, special software is used to allocate and manage the execution of the parallel tasks of the applications on clusters processors to satisfy real-time requirements and achieve high throughput. In this paper, we present a new algorithm to allocate tasks’ workloads in a way that improves the utilization of each clusters processor. Consequently, a clusters processor can accept more tasks and results in higher throughput. The idea of the algorithm depends on assigning a variable processing power to the task under consideration to satisfy its deadline instead of rejecting it if a constant processing power cannot be guaranteed. Simulation results reveal that the acceptance rate of submitted tasks to a clusters processor using the new approach is superior to previous approaches.

Textual Signs Reading for Indoor Semantic Map Construction

Semantic maps have come up to facilitate high-level human robot interaction like “deliver a laptop for maintenance” . Textual signs posted on the walls and room doors in indoor environments are environment features that could be easily identified by service robots. These signs provide some semantic information like the function of the room and its occupants. In addition, they consider landmarks for the service robots. This paper is a continuous work of developing a framework for creating a semantic map for indoor environment using mobile robot [17]. The paper addresses the problem of automatically detecting and recognizing textual signs during robotic mapping. Then, annotating these recognized signs to a previously generated robotic grid map. ATRV-mini robot has been used in our experiments.

Efficient processing power utilization to execute real-time tasks

Scheduling real-time applications requires high utilization of the available processing power on processors to accommodate as many tasks as possible while satisfying the required deadlines of the applications. Moreover, an acceptable level of reliability should be provided due to the critical nature of the applications. In this paper, we propose a new approach that highly utilizes the available processing power to improve the admission rate of real-time task scheduling. The basic idea is to use all of the available processing power to execute allocated tasks. This reduces the execution time needed for each task and hence increases its reliability. It also reduces needed processing power for existing tasks to satisfy their deadlines and hence provides higher available processing power for any new task. This gives a higher chance for the new arriving task to be accommodated. We show that our approach is beneficial in that it provides a better scheme with higher acceptance rate for real time tasks compared to the traditional approach of using a fixed processing power reservation

Topic modelling for automatic selection of software design patterns

Design pattern is a high-quality and reusable solution to a recurring software design problem. It is considered an important concept in the software engineering field due to its ability to enhance some of the quality attributes of the software systems including maintainability and extensibility. However, novice developers need to be provided by a tool to assist them in selecting the fit design pattern to solve a design problem. The paper proposes a novel approach for the automatic selection of the fit design pattern. This approach is based on using Latent Dirichlet Allocation (LDA) topic model. The topic is a set of words that often appear together. LDA is able to relate words with similar meaning and to differentiate between uses of words with multiple meanings. In this paper LDA is used to analyze the textual descriptions of design patterns and extract the topics then discover the similarity between the target problem scenario and the collection of patterns using Improved Sqrt-Cosine similarity measure (ISCS). The proposed approach was evaluated using Gang of four design patterns. The experimental results showed that the proposed approach outperforms approach based on the traditional vector space model of Unigrams.

Automatic Recommendation of Software Design Patterns: Text Retrieval Approach

Design pattern is a reusable solution to a commonly occurring design problem in certain context. Using design patterns in software development improves the product’s quality, understandability and productivity. However, it is a challenging task for novice developers to select the right design pattern to solve a design problem. The paper proposes a methodology for the automatic selection of the fit design pattern from a list of patterns. The proposed methodology is based on Text retrieval approach where the design problem scenarios are described in natural language. A vector space model (VSM) was created for the catalogue of design patterns. A vector of features consists of unigrams and bigrams is generated for the given design problem scenario. The recommended design pattern is the closest to the problem scenario. The proposed mechanism was evaluated using the Gang of four design patterns and the experimental results showed the effectiveness of the proposed methodology.

Apparel coordination based on contour and dominant colours matching

The paper proposes a mobile application for clothing coordination, which could be of great benefit for stores and people seek for fashion advices. The application matches apparel image input with, previously saved apparel images, and then provides the user with the possible matching suggestions based on the apparel outline and dominating colors. For this purpose two Region of Interest (ROI) extraction components were developed, to facilitate apparel outline classification and color extraction. Two-level binary support vector machines were trained for apparels classification into top, skirt or pair of trousers. K-means algorithm was used for recognizing the apparels dominant colors. The colour matching is offered through analogous and complementary matching; K-nearest neighbor algorithm was adopted for this purpose. The application was developed and tested on Android 2.2 (Froyo) and the results showed that: (1) the application is able to classify apparels at high accuracy, (2) The proposed color recognition and matching strategies outperform the current ones.

Android Application to Assist Visually Impaired with Outfit Coordination

ABSTRACT The paper presents an android based application to assist visually impaired with outfit coordination process. The application helps them to be independent in their decisions while shopping or getting ready for the day. The application matches apparel image input with users previously saved closet items, then provides the user with the possible matching item suggestions. The matching process is done based on the item outline and dominating colours. In order to develop this application five main components have been developed: 1. Two through analogous and complementary matching, as opposed to Region of interest (ROI) extraction components facilitating feature and colour extraction; 2. An outline detection component to determine whether an item is a top, skirt or a pair of trousers; 3. A colour recognition component to extract dominant colours of an apparel image; 4. A descriptive colour verbal feedback; 5. A matching component based on item outline and colours to coordinate with others in the user‟s closet. The output to the user includes the descriptive colours, communicated via audio, indicating the 3 dominant colours using their descriptive names - selected from a pool of 581 different colour shades. Other output is the outline details of the matching items. The application was developed and tested on Android 2.2 (Froyo) and the results showed that the ROI extraction and outline detection components perform well. The colour recognition and descriptive name generation modules outperforms the current ones.

Real-Time Workload Allocation on a Uni-processor

The paper presents a novel allocation algorithm to allocate independent real time tasks on a processor in a way that improves the processor’s throughput (Processor’s throughput is the number of tasks the processor can accept for execution). The proposed approach allocates tasks’ workloads (task’s workload is the percentage of work required by the processor to execute the task) instead of their processing powers (Processing power assigned to a task is a percentage of the processor reserved to execute the task such that its deadline is satisfied). To achieve our objective a variable processing power is assigned to the task under consideration over its deadline to satisfy its timing requirements instead of rejecting it if a constant processing power cannot be guaranteed as in previous CPU reservation approaches. Simulation results revealed that the acceptance rate of the admitted tasks to a certain processor using the new approach is superior to that achieved using the traditional processing power reservation approach.

Efficient Execution of Real-Time Tasks on a Single Process

This paper presents a scheduling algorithm that maximizes utilization of available processing power and consequently maximizes the throughput of a single processor. The main idea is to distribute all available processing power over the tasks being executed proportional to their workload such that each task can get at least its reserved processing power. If a task received a processing power more than the required value, its execution time reduces and it finishes before its deadline. When a new task arrives, a processors processing power is re-distributed among the tasks being executed on this processor to complete their jobs without violating their deadlines. Consequently, a higher workload becomes available for the new arrived task. This gives a higher chance for the new task to be admitted. If the available processing power of the target processor is not enough to accept the new task, but there is enough available workload, a variable processing power is assigned to the arrived task during its execution to allocate the task to the processor without violating its deadline. Experimental results reveal that the proposed approach improves acceptance rate compared to three previous approaches