Steered Beam Adaptive Antenna Arrays

In this chapter, the performance of steered beam adaptive arrays is presented with its corresponding analytical expressions. Computer simulations are used to illustrate the performance of the array under various operating conditions. In this chapter, we ignore the presence of mutual coupling between the array elements. The principal system elements of the adaptive array consist of an array of sensors (antennas), a pattern-forming network, and an adaptive pattern control unit or adaptive processor that adjusts the variable weights in the pattern-forming network. The adaptive pattern control unit may furthermore be conveniently subdivided into a signal processor unit and an adaptive control algorithm. The manner in which these elements are actually implemented depends on the propagation medium in which the array is to operate, the frequency spectrum of interest, and the user’s knowledge of the operational signal environment.

Room-temperature control and electrical readout of individual nitrogen-vacancy nuclear spins

Nuclear spins in semiconductors are leading candidates for future quantum technologies, including quantum computation, communication, and sensing. Nuclear spins in diamond are particularly attractive due to their long coherence time. With the nitrogen-vacancy (NV) centre, such nuclear qubits benefit from an auxiliary electronic qubit, which, at cryogenic temperatures, enables probabilistic entanglement mediated optically by photonic links. Here, we demonstrate a concept of a microelectronic quantum device at ambient conditions using diamond as wide bandgap semiconductor. The basic quantum processor unit – a single 14N nuclear spin coupled to the NV electron – is read photoelectrically and thus operates in a manner compatible with nanoscale electronics. The underlying theory provides the key ingredients for photoelectric quantum gate operations and readout of nuclear qubit registers. This demonstration is, therefore, a step towards diamond quantum devices with a readout area limited by inter-electrode distance rather than by the diffraction limit. Such scalability could enable the development of electronic quantum processors based on the dipolar interaction of spin-qubits placed at nanoscopic proximity.

Smart Video Surveillance System Based on Edge Computing

New processing methods based on artificial intelligence (AI) and deep learning are replacing traditional computer vision algorithms. The more advanced systems can process huge amounts of data in large computing facilities. In contrast, this paper presents a smart video surveillance system executing AI algorithms in low power consumption embedded devices. The computer vision algorithm, typical for surveillance applications, aims to detect, count and track people’s movements in the area. This application requires a distributed smart camera system. The proposed AI application allows detecting people in the surveillance area using a MobileNet-SSD architecture. In addition, using a robust Kalman filter bank, the algorithm can keep track of people in the video also providing people counting information. The detection results are excellent considering the constraints imposed on the process. The selected architecture for the edge node is based on a UpSquared2 device that includes a vision processor unit (VPU) capable of accelerating the AI CNN inference. The results section provides information about the image processing time when multiple video cameras are connected to the same edge node, people detection precision and recall curves, and the energy consumption of the system. The discussion of results shows the usefulness of deploying this smart camera node throughout a distributed surveillance system.

Estimating expectation values using approximate quantum states

We introduce an approximate description of an N-qubit state, which contains sufficient information to estimate the expectation value of any observable to a precision that is upper bounded by the ratio of a suitably-defined seminorm of the observable to the square root of the number of the system's identical preparations M, with no explicit dependence on N. We describe an operational procedure for constructing the approximate description of the state that requires, besides the quantum state preparation, only single-qubit rotations followed by single-qubit measurements. We show that following this procedure, the cardinality of the resulting description of the state grows as 3MN. We test the proposed method on Rigetti's quantum processor unit with 12, 16 and 25 qubits for random states and random observables, and find an excellent agreement with the theory, despite experimental errors.

Real Time Generation of Three Dimensional Patterns for Multiphoton Stimulation

The advent of optogenetics has revolutionized experimental research in the field of Neuroscience and the possibility to selectively stimulate neurons in 3D volumes has opened new routes in the understanding of brain dynamics and functions. The combination of multiphoton excitation and optogenetic methods allows to identify and excite specific neuronal targets by means of the generation of cloud of excitation points. The most widely employed approach to produce the points cloud is through a spatial light modulation (SLM) which works with a refresh rate of tens of Hz. However, the computational time requested to calculate 3D patterns ranges between a few seconds and a few minutes, strongly limiting the overall performance of the system. The maximum speed of SLM can in fact be employed either with high quality patterns embedded into pre-calculated sequences or with low quality patterns for real time update. Here, we propose the implementation of a recently developed compressed sensing Gerchberg-Saxton algorithm on a consumer graphical processor unit allowing the generation of high quality patterns at video rate. This, would in turn dramatically reduce dead times in the experimental sessions, and could enable applications previously impossible, such as the control of neuronal network activity driven by the feedback from single neurons functional signals detected through calcium or voltage imaging or the real time compensation of motion artifacts.

Implementation of Autoencoders with Systolic Arrays through OpenCL

In the world of algorithm acceleration and the implementation of deep neural networks’ recall phase, OpenCL based solutions have a clear tendency to produce perfectly adapted kernels in graphic processor unit (GPU) architectures. However, they fail to obtain the same results when applied to field-programmable gate array (FPGA) based architectures. This situation, along with an enormous advance in new GPU architectures, makes it unfeasible to defend an acceleration solution based on FPGA, even in terms of energy efficiency. Our goal in this paper is to demonstrate that multikernel structures can be written based on classic systolic arrays in OpenCL, trying to extract the most advanced features of FPGAs without having to resort to traditional FPGA development using lower level hardware description languages (HDLs) such as Verilog or VHDL. This OpenCL methodology is based on the intensive use of channels (IntelFPGA extension of OpenCL) for the communication of both data and control and on the refinement of the OpenCL libraries using register transfer logic (RTL) code to improve the performance of the implementation of the base and activation functions of the neurons and, above all, to reflect the importance of adequate communication between the layers when implementing neuronal networks.

Review on Sparse Matrix Storage Formats With Space Complexity Analysis

Sparse matrix-vector multiplication (SpMV) is a challenging computational kernel in linear algebra applications, like data mining, image processing, and machine learning. The performance of this kernel is greatly dependent on the size of the input matrix and the underlying hardware features. Various sparse matrix storage formats referred to commonly as sparse formats have been proposed in the literature to reduce the size of the matrix. In modern multi-core and many-core architectures, the performance of the kernel is mainly dependent on memory wall and power wall problem. Normally review on sparse formats is done with specific architecture or with specific application. This chapter presents a comparative study on various sparse formats in cross platform architecture like CPU, graphics processor unit (GPU), and single instruction multiple data stream (SIMD) registers. Space complexity analysis of various formats with its representation is discussed. Finally, the merits and demerits of each format have been summarized into a table.

Quantum Approximated Graph Cutting: A Rapid Replacement for T-REMD?

Determining an optimal protein configuration for the employment of protein binding analysis as completed by Temperature based Replica Exchange Molecular Dynamics (T-REMD) is an important process in the accurate depiction of a protein's behavior in different solvent environments, especially when determining a protein's top binding sites for use in protein-ligand and protein-protein docking studies. However, the completion of this analysis, which pushes out top binding sites through configurational changes, is an polynomial-state computational problem that can take multiple hours to compute, even on the fastest supercomputers. In this study, we aim to determine if graph cutting provide approximated solutions the MaxCut problem can be used as a method to provide similar results to T-REMD in the determination of top binding sites of Surfactant Protein A (SP-A) for binding analysis. Additionally, we utilize a quantum-hybrid algorithm within Iff Technologies' Polar+ package using an actual quantum processor unit (QPU), an implementation of Polar+ using an emulated QPU, or Quantum Abstract Machine (QAM), on a large scale classical computing device, and an implementation of a classical MaxCut algorithm on a supercomputer in order to determine the types of advantages that can be gained through using a quantum computing device for this problem, or even using quantum algorithms on a classical device. This study found that Polar+ provides a dramatic speedup over a classical implementation of a MaxCut approximation algorithm or the use of GROMACS T-REMD, and produces viable results, in both its QPU and QAM implementations. However, the lack of direct configurational changes carried out onto the structure of SP-A after the use of graph cutting methods produces different final binding results than those produced by GROMACS T-REMD. Thus, further work needs to be completed into translating quantum-based probabilities into configurational changes based on a variety of noise conditions to better determine the accuracy advantage that quantum algorithms and quantum devices can provide in the near future.