Elankovan Sundararajan

Review: Cloud computing service composition: A systematic literature review

The increasing tendency of network service users to use cloud computing encourages web service vendors to supply services that have different functional and nonfunctional (quality of service) features and provide them in a service pool. Based on supply and demand rules and because of the exuberant growth of the services that are offered, cloud service brokers face tough competition against each other in providing quality of service enhancements. Such competition leads to a difficult and complicated process to provide simple service selection and composition in supplying composite services in the cloud, which should be considered an NP-hard problem. How to select appropriate services from the service pool, overcome composition restrictions, determine the importance of different quality of service parameters, focus on the dynamic characteristics of the problem, and address rapid changes in the properties of the services and network appear to be among the most important issues that must be investigated and addressed. In this paper, utilizing a systematic literature review, important questions that can be raised about the research performed in addressing the above-mentioned problem have been extracted and put forth. Then, by dividing the research into four main groups based on the problem-solving approaches and identifying the investigated quality of service parameters, intended objectives, and developing environments, beneficial results and statistics are obtained that can contribute to future research.

Performance study of selective encryption in comparison to full encryption for still visual images

Securing digital images is becoming an important concern in today’s information security due to the extensive use of secure images that are either transmitted over a network or stored on disks. Image encryption is the most effective way to fulfil confidentiality and protect the privacy of images. Nevertheless, owing to the large size and complex structure of digital images, the computational overhead and processing time needed to carry out full image encryption prove to be limiting factors that inhibit it of being used more heavily in real time. To solve this problem, many recent studies use the selective encryption approach to encrypt significant parts of images with a hope to reduce the encryption overhead. However, it is necessary to realistically evaluate its performance compared to full encryption. In this paper, we study the performance and efficiency of image segmentation methods used in the selective encryption approach, such as edges and face detection methods, in determining the most important parts of visual images. Experiments were performed to analyse the computational results obtained by selective image encryption compared to full image encryption using symmetric encryption algorithms. Experiment results have proven that the selective encryption approach based on edge and face detection can significantly reduce the time of encrypting still visual images as compared to full encryption. Thus, this approach can be considered a good alternative in the implementation of real-time applications that require adequate security levels.

A hybrid imperialist competitive-gravitational attraction search algorithm to optimize cloud service composition

Service composition is among the most important challenges that cloud providers have ever faced. Optimization of QoS attributes when composing simple atomic services to obtain a complex service can be considered to be an NP-hard problem, which could be solved properly by using Hybrid optimization algorithms. In this research, the hybridization of an improved Gravitational Attraction Search (as a local search algorithm) with an Imperialist Competitive Algorithm has led us to introduce and apply a new memetic algorithm for gaining optimal or near optimal response time and execution fees simultaneously, for cloud computing service composition. Using a roulette wheel selection algorithm to make well-advised and non-blind decisions to choose the number of countries in each empire that should be selected to apply a local search to has assisted the hybrid algorithm at achieving better solutions. Introducing a new equation to calculate the QoS eligibility of the solutions that were generated based on the normalization of the response time and execution fee has also led us to compute the results fairly and in a scientifically based manner.

Fitted dynamic framed slotted ALOHA anti-collision algorithm in RFID systems

Radio Frequency Identification (RFID) system is a new communication technology to identify the objects using electromagnetic waves. The main superiority of RFID systems in comparison with other identification systems, such as barcodes, is its ability in simultaneous identification process, but when more than one tag wants to communicate with the reader, collision problem occurs and wastes time and increases energy consumption. This reduces efficiency of the identification process in RFID systems. Collision problem happens when the reader has to identify more than one tag at the same time. Consequently, it is essential to develop an efficient anti-collision algorithm to save the time and energy and increase the performance of the system. In this paper, after investigating the previous works on ALOHA-based anti-collision algorithms, a new anti-collision algorithm is presented to improve the DFSA ALOHA-based anti-collision algorithms. Evaluation of the ALOHA-based algorithms and simulation of the results illustrate that presented work recognizes the tags more efficiently by reducing the number of time slots.

Cubic spline-based tag estimation method in RFID multi-tags identification process

Radio Frequency Identification (RFID) system is a communication technology used to identify objects using electromagnetic waves. The key advantage of RFID systems stems from their ability to simultaneously identify multiple tagged objects. However, communication of multiple tags with a reader may result in a collision problem, which is both time and energy inefficient, hindering the effectiveness of tag identification process. Presently, several anti-collision algorithms can be applied in order to reduce the collision probability. The reader¿s a priori knowledge of tag quantity significantly affects the overall performance of the system. Since the exact number of tags is not available for the reader, it is essential to develop an accurate tag estimation method to increase the efficiency of tag identification process. This paper presents a novel tag quantity estimation method, whereby, after simulating the tag distribution process, cubic spline interpolation method is employed to approximate the number of tags. According to the simulation results and the evaluation of the previous estimation methods, the new proposed method estimates the number of tags with a higher accuracy yielding an error rate of less than 1%, on average. Moreover, this low error rate is preserved even when the number of tags increases considerably.

A parallel membrane inspired harmony search for optimization problems

Display Omitted Harmony Search is enhanced by improving the speed of convergence while preventing premature convergence to a local minimum.Membrane computing model is employed to execute parallelized HS efficiently to increase the diversity of HS and improving the performance of HS.Simulation results showed that the application of the proposed approach to different variants of harmony search yielded better performance than previous approaches.We applied the parallel membrane-inspired HS to the flexible job shop scheduling problem and well-known benchmark instances to demonstrate the effectiveness of the proposed algorithm. Harmony search is an emerging meta-heuristic optimization algorithm that is inspired by musical improvisation processes, and it can solve various optimization problems. Membrane computing is a distributed and parallel model for solving hard optimization problems. First, we employed some previously proposed approaches to improve standard harmony search by allowing its parameters to be adaptive during the processing steps. Information from the best solutions was used to improve the speed of convergence while preventing premature convergence to a local minimum. Second, we introduced a parallel framework based on membrane computing to improve the harmony search. Our approach utilized the parallel membrane computing model to execute parallelized harmony search efficiently on different cores, where the membrane computing communication characteristics were used to exchange information between the solutions on different cores, thereby increasing the diversity of harmony search and improving the performance of harmony search. Our simulation results showed that the application of the proposed approach to different variants of harmony search yielded better performance than previous approaches. Furthermore, we applied the parallel membrane inspired harmony search to the flexible job shop scheduling problem. Experiments using well-known benchmark instances showed the effectiveness of the algorithm.

Parallel and distributed computing models on a graphics processing unit to accelerate simulation of membrane systems

Abstract Membrane systems are parallel distributed computing models that are used in a wide variety of areas. Use of a sequential machine to simulate membrane systems loses the advantage of parallelism in Membrane Computing. In this paper, an innovative classification algorithm based on a weighted network is introduced. Two new algorithms have been proposed for simulating membrane systems models on a Graphics Processing Unit (GPU). Communication and synchronization between threads and thread blocks in a GPU are time-consuming processes. In previous studies, dependent objects were assigned to different threads. This increases the need for communication between threads, and as a result, performance decreases. In previous studies, dependent membranes have also been assigned to different thread blocks, requiring inter-block communications and decreasing performance. The speedup of the proposed algorithm on a GPU that classifies dependent objects using a sequential approach, for example with 512 objects per membrane, was 82×, while for the previous approach ( Algorithm 1 ), it was 8.2×. For a membrane system with high dependency among membranes, the speedup of the second proposed algorithm ( Algorithm 3 ) was 12×, while for the previous approach ( Algorithm 1 ) and the first proposed algorithm ( Algorithm 2 ) that assign each membrane to one thread block, it was 1.8×.

Extending birthday paradox theory to estimate the number of tags in RFID systems.

The main objective of Radio Frequency Identification systems is to provide fast identification for tagged objects. However, there is always a chance of collision, when tags transmit their data to the reader simultaneously. Collision is a time-consuming event that reduces the performance of RFID systems. Consequently, several anti-collision algorithms have been proposed in the literature. Dynamic Framed Slotted ALOHA (DFSA) is one of the most popular of these algorithms. DFSA dynamically modifies the frame size based on the number of tags. Since the real number of tags is unknown, it needs to be estimated. Therefore, an accurate tag estimation method has an important role in increasing the efficiency and overall performance of the tag identification process. In this paper, we propose a novel estimation technique for DFSA anti-collision algorithms that applies birthday paradox theory to estimate the number of tags accurately. The analytical discussion and simulation results prove that the proposed method increases the accuracy of tag estimation and, consequently, outperforms previous schemes.

Incorporating Fault Tolerance with Replication on Very Large Scale Grids

Providing fault tolerance for message passing parallel application on a distributed environment is a rule rather than an exception. A node failure can cause the whole computation to stop and has to be restarted from the beginning if no fault tolerance is available. However, introducing fault tolerance has some overhead on speedup that can be achieved. In this paper, we introduce a new technique called replication with cross-over packets for reliability and to increase fault tolerance over Very Large Scale Grids (VLSG). This technique has two pronged effect of avoiding single point of failure and single link of failure. We incorporate this new technique into the L-BSP model and show the possible speedup of parallel process. We also derive the achievable speedup for some fundamental parallel algorithms using this technique.

Using computer labs for distributed computing

PCs in computing labs at institution of higher learning are often if not always underutilized. These PCs are usually not used at night and early morning (i.e. between 10pm–8am), during semester breaks, weekends and during final examinations. At other times most of the lab PCs are not utilized up to the available capacity. This underutilization scenario provides an excellent opportunity for researchers to utilize these lab PCs for research work that requires immense computing power. Furthermore, it is noteworthy that the life cycle of a PC is about 5 years which comes to about 1826 days (1 year = 365.25 days). At the end of its life cycle, new PCs are often purchased to replace the older ones. Thus, using these PCs effectively before being rendered obsolete should be given high priority. The use of distributed computing for compute intensive problems is not new. However, with the availability of improved hardware, fast interconnects and user friendly middleware, the use of distributed computing has become even more popular now. In this paper, we discuss our experience in using Condor middleware on lab PCs at Faculty of Information Science and Technology. We have recorded good speedups from our experiments for embarrassingly parallel problem that we executed on the lab PCs. A script and programs were written to automate the process of preparing input files, merging output files, and displaying output. The single command execution provided by the script offers user friendly environment for the users. We hope the outcome of this experiment will encourage more faculty members to use the available resources at the computing labs for their research work.