S. M. Farhad

Profile-guided deployment of stream programs on multicores

Because multicore architectures have become the industry standard, programming abstractions for concurrent programming are of key importance. Stream programming languages facilitate application domains characterized by regular sequences of data, such as multimedia, graphics, signal processing and networking. With stream programs, computations are expressed through independent actors that interact through FIFO data channels. A major challenge with stream programs is to load-balance actors among available processing cores. The workload of a stream program is determined by actor execution times and the communication overhead induced by data channels. Estimating communication costs on cache-coherent shared-memory multiprocessors is difficult, because data movements are abstracted away by the cache coherence protocol. Standard execution time profiling techniques cannot separate actor execution times from communication costs, because communication costs manifest in terms of execution time overhead. In this work we present a unified Integer Linear Programming (ILP) formulation that balances the workload of stream programs on cache-coherent multicore architectures. For estimating the communication costs of data channels, we devise a novel profiling scheme that minimizes the number of profiling steps. We conduct experiments across a range of StreamIt benchmarks and show that our method achieves a speedup of up to 4.02x on 6 processors. The number of profiling steps is on average only 17% of an exhaustive profiling run over all data channels of a stream program.

Fingerprint Recognition by Chain Coded String Matching Technique

Research in automated fingerprint recognition has been carried out for decades. But reliable and faster fingerprint recognition is still an open problem. Determination of features, encoding them and matching with template are the most challenging problems faced in this area. This paper introduces a novel approach of fingerprint recognition based on Flow Pattern and Chain Coded String Matching technique. In our technique, we propose to use string encoding of images from a Circular Tessellation for recognizing fingerprint data. This approach has a significant improvement in time complexity over the typical square tessellated and gray scale representation techniques.

Internet of Things Based Free Parking Space Management System

Internet of things is an emerging technology that enables every object of our surroundings to be connected and controlled from a distance. Wireless sensors can be integrated to the internet and the sensor data can be monitored and thus detect events from distance. Reliable free parking-space management system is important to maintain the discipline and better utilization of the parking space. Free parkingspace management system using wireless sensor networks is less expensive and requires less installation requirements. In this paper, we have illustrated a complete implementation of free parking-space management system using wireless sensor networks. We have used sensor motes of Texas Instruments each consisting of ten sensors. We have employed four sensors, magnetic, light, temperature, and ambient temperature sensors of the motes for reliable detection of the free space. We have optimized the operating system, adopted low-powered efficient networking protocol and a lightweight messaging model for data transmission. We have implemented the system and deployed in a parking area. Our experimental results show that the system can detect vehicle presence with an accuracy of 98%.

The optimal window size for analysing longitudinal networks

The time interval between two snapshots is referred to as the window size. A given longitudinal network can be analysed from various actor-level perspectives, such as exploring how actors change their degree centrality values or participation statistics over time. Determining the optimal window size for the analysis of a given longitudinal network from different actor-level perspectives is a well-researched network science problem. Many researchers have attempted to develop a solution to this problem by considering different approaches; however, to date, no comprehensive and well-acknowledged solution that can be applied to various longitudinal networks has been found. We propose a novel approach to this problem that involves determining the correct window size when a given longitudinal network is analysed from different actor-level perspectives. The approach is based on the concept of actor-level dynamicity, which captures variability in the structural behaviours of actors in a given longitudinal network. The approach is applied to four real-world, variable-sized longitudinal networks to determine their optimal window sizes. The optimal window length for each network, determined using the approach proposed in this paper, is further evaluated via time series and data mining methods to validate its optimality. Implications of this approach are discussed in this article.

Temperature Sensitive Microarchitecture Design Circuit Design

Microprocessors are designed with very tiny microchips and heat induced due to operation makes the chip deteriorate their performance in many extents. Heat causes a portion of chiparea to get beyond tolerable temperature range which can degrade performance of many applications in chip-level. This work addresses many issues in this range. The main contribution of the work lies in reconsidering the heat transition among chips, or inside a chip in order to decrease heat inside a microprocessor. With this end in view, a renewed design architecture in circuit-level has been considered. To design the total work inside microprocessor in response to dynamic temperature change a different level of operation-mechanism has been proposed. To watch the applications running in pipeline, and then by utilizing slack time in hardware level this work wants to improve performance of the processor. In brief, total work proposes two new heat-control mechanisms, one is at operation-level and the other is at architectural-level. At operation-level, this work proposes a prediction mechanism to predict the useful operations inside the microprocessor that performs as a sink for heat dissipation. At architectural-level, this work proposes a drain system. This work has simulated the proposed system using Matlab and observed that the system works perfectly well. A comparison with existing mechanism has been devised which shows the proposed work increases performance of running application.

Dynamic Resource Provisioning for Video Transcoding in IaaS Cloud

High-definition video applications are often challenging for mobile devices due to their limited processing capability and bandwidth-constrained network connection. Video transcoding has become an inevitable technology for on-demand video streaming service which needs to be done on the go in real-time for mobile devices. Since video transcoding involves extensive computation, performing transcoding using cloud resource is more cost friendly and time effective. It is challenging to use the cloud resources for video transcoding that minimises the operating cost. In this paper, we propose a dynamic resource provisioning algorithm for allocating virtual machine to scale video transcoding services on a given IaaS cloud and conducted experiments in CloudSim and compared the proposed system with conventional system. The experimental results show that the proposed system outperforms the conventional systems.