An Enhanced Electroporator Design for Pulse Generation

An electroporator is an instrument used for delivering electrical pulses to a tumor. In this work, an electroporator consisting of three main system blocks, namely High Voltage (HV) source, nanosecond (ns) switching, and pulse generation, was designed, developed, and evaluated to generate high voltage ns pulses to treat tumors. The high-voltage source block was used to convert the 9.6V DC from the battery to a variable HV output and store this charge for later use. The ns switching block contained a MOSFET-based low-side switch which applies short ns pulses to the load. The pulse generation block generates short ns pulses and supplies the adequate current to turn on the MOSFET at a quicker rate aiding the application of these pulses to the load. This process was simulated using PSpice software and the results are presented.

ANALISIS PERBANDINGAN PELACAKAN OBJEK MENGGUNAKAN ALGORITMA HORN-SCHUNCK DAN LUCAS-KANADE

Object tracking one of computer vision. Computer vision similar to human eye function. The difficulty is to detect presence an object and object tracking application made. Object tracking used in aircraft, track cars, human body detectors at airports, a regulator the number of vehicles pass and navigation tools on robots. This study is to identify objects that pass in frame. This research also count the number of objects that pass in one frame. Object tracking done by comparing two algorithms namely Horn-Schunck and Lucas-Kanade. Both algorithms tested using the Source Block Parameter and Function Block Parameter. The test carried out with video resolution 120x160 and the position camera is 2-4 m. The object tracking test is conducted in the duration of 110-120 seconds. Stages tracking object was thresholding, filtering and region successfully obtain object binary video. The Lucas-Kanade has faster in identifying objects compared to the Horn-Schunck algorithm.

CFD Simulation of Natural Convection Flow and Heat Transfer Process in Rectangular Cavity

In this paper we investigate the natural convective heat transfer process inside a ventilated rectangular cavity with a projected heat source. The heat source block is mounted on the bottom wall and a horizontal vent is provided on the top wall of the rectangular cavity. The flow is induced due to the density difference which arises due to the variations in temperature between the heat source block and the surrounding ambient fluid. A FORTRAN 90 CFD solver is developed to simulate the natural convection phenomena by solving the Navier-stokes equation, energy equation coupled with Realizable k-ε turbulence model. The transient flow behavior inside the cavity is simulated by varying the heat source aspect ratios, Grashof number and the heat source locations. It is found that the heat source aspect ratio and its locations significantly influences the flow and heat transfer characteristics inside the cavity. The bidirectional exchange rate across the horizontal opening increases linearly with Grashof number and heat source aspect ratio. A chaotic flow behavior pattern is observed across the opening and the strength of the instabilities increases linearly with heat source aspect ratio. It is identified that by varying the aspect ratio 0.1 ≤ β ≤ 3, the average Nusselt number and mass flow rates are increased by 28% and 43% respectively.

Hadoop distributed file system mechanism for processing of large datasets across computers cluster using programming techniques

In this paper, we have proved that the HDFS I/O operations performance is getting increased by integrating the set associativity in the cache design and changing the pipeline topology using fully connected digraph network topology. In read operation, since there is huge number of locations (words) at cache compared to direct mapping the chances of miss ratio is very low, hence reducing the swapping of the data between main memory and cache memory. This is increasing the memory I/O operations performance. In Write operation instead of using the sequential pipeline we need to construct the fully connected graph using the data blocks listed from the NameNode metadata. In sequential pipeline, the data is getting copied to source node in the pipeline. Source node will copy the data to next data block in the pipeline. The same copy process will continue until the last data block in the pipeline. The acknowledgment process has to follow the same process from last block to source block. The time required to transfer the data to all the data blocks in the pipeline and the acknowledgment process is almost 2n times to data copy time from one data block to another data block (if the replication factor is n).

Network Recasting: A Universal Method for Network Architecture Transformation

This paper proposes network recasting as a general method for network architecture transformation. The primary goal of this method is to accelerate the inference process through the transformation, but there can be many other practical applications. The method is based on block-wise recasting; it recasts each source block in a pre-trained teacher network to a target block in a student network. For the recasting, a target block is trained such that its output activation approximates that of the source block. Such a block-by-block recasting in a sequential manner transforms the network architecture while preserving the accuracy. This method can be used to transform an arbitrary teacher network type to an arbitrary student network type. It can even generate a mixed-architecture network that consists of two or more types of block. The network recasting can generate a network with fewer parameters and/or activations, which reduce the inference time significantly. Naturally, it can be used for network compression by recasting a trained network into a smaller network of the same type. Our experiments show that it outperforms previous compression approaches in terms of actual speedup on a GPU.

Exponential Strong Converse for One Helper Source Coding Problem

We consider the one helper source coding problem posed and investigated by Ahlswede, Körner and Wyner. Two correlated sources are separately encoded and are sent to a destination where the decoder wishes to decode one of the two sources with an arbitrary small error probability of decoding. In this system, the error probability of decoding goes to one as the source block length n goes to infinity. This implies that we have a strong converse theorem for the one helper source coding problem. In this paper, we provide the much stronger version of this strong converse theorem for the one helper source coding problem. We prove that the error probability of decoding tends to one exponentially and derive an explicit lower bound of this exponent function.

Efficient Quantization with Linear Index Coding for Deep-Space Images

Due to inevitable propagation delay involved in deep-space communication systems, very high cost is associated with the retransmission of erroneous segments. Quantization with linear index coding (QLIC) scheme is known to provide compression along with robust transmission of deep-space images, and thus the likelihood of retransmissions is significantly reduced. This paper aims to improve its spectral efficiency as well as robustness. First, multiple quantization refinement levels per transmitted source block of QLIC are proposed to increase spectral efficiency. Then, iterative multipass decoding is introduced to jointly decode the subsource symbol-planes. It achieves better PSNR of the reconstructed image as compared to the baseline one-pass decoding approach of QLIC.