Memory hierarchy and data communication in heterogeneous reconfigurable SoCs

The miniaturization race in the hardware industry aiming at continuous increasing of transistor density on a die does not bring respective application performance improvements any more. One of the most promising alternatives is to exploit a heterogeneous nature of common applications in hardware. Supported by reconfigurable computation, which has already proved its efficiency in accelerating data intensive applications, this concept promises a breakthrough in contemporary technology development. Memory organization in such heterogeneous reconfigurable architectures becomes very critical. Two primary aspects introduce a sophisticated trade-off. On the one hand, a memory subsystem should provide well organized distributed data structure and guarantee the required data bandwidth. On the other hand, it should hide the heterogeneous hardware structure from the end-user, in order to support feasible high-level programmability of the system. This thesis work explores the heterogeneous reconfigurable hardware architectures and presents possible solutions to cope the problem of memory organization and data structure. By the example of the MORPHEUS heterogeneous platform, the discussion follows the complete design cycle, starting from decision making and justification, until hardware realization. Particular emphasis is made on the methods to support high system performance, meet application requirements, and provide a user-friendly programmer interface. As a result, the research introduces a complete heterogeneous platform enhanced with a hierarchical memory organization, which copes with its task by means of separating computation from communication, providing reconfigurable engines with computation and configuration data, and unification of heterogeneous computational devices using local storage buffers. It is distinguished from the related solutions by distributed data-flow organization, specifically engineered mechanisms to operate with data on local domains, particular communication infrastructure based on Network-on-Chip, and thorough methods to prevent computation and communication stalls. In addition, a novel advanced technique to accelerate memory access was developed and implemented.

LBT AO system: tests and calibration

An Adaptive Optic (AO) system is a fundamental requirement of 8m-class telescopes. We know that in order to obtain the maximum possible resolution allowed by these telescopes we need to correct the atmospheric turbulence. Thanks to adaptive optic systems we are able to use all the effective potential of these instruments, drawing all the information from the universe sources as best as possible. In an AO system there are two main components: the wavefront sensor (WFS) that is able to measure the aberrations on the incoming wavefront in the telescope, and the deformable mirror (DM) that is able to assume a shape opposite to the one measured by the sensor. The two subsystem are connected by the reconstructor (REC). In order to do this, the REC requires a “common language" between these two main AO components. It means that it needs a mapping between the sensor-space and the mirror-space, called an interaction matrix (IM). Therefore, in order to operate correctly, an AO system has a main requirement: the measure of an IM in order to obtain a calibration of the whole AO system. The IM measurement is a 'mile stone' for an AO system and must be done regardless of the telescope size or class. Usually, this calibration step is done adding to the telescope system an auxiliary artificial source of light (i.e a fiber) that illuminates both the deformable mirror and the sensor, permitting the calibration of the AO system. For large telescope (more than 8m, like Extremely Large Telescopes, ELTs) the fiber based IM measurement requires challenging optical setups that in some cases are also impractical to build. In these cases, new techniques to measure the IM are needed. In this PhD work we want to check the possibility of a different method of calibration that can be applied directly on sky, at the telescope, without any auxiliary source. Such a technique can be used to calibrate AO system on a telescope of any size. We want to test the new calibration technique, called “sinusoidal modulation technique”, on the Large Binocular Telescope (LBT) AO system, which is already a complete AO system with the two main components: a secondary deformable mirror with by 672 actuators, and a pyramid wavefront sensor. My first phase of PhD work was helping to implement the WFS board (containing the pyramid sensor and all the auxiliary optical components) working both optical alignments and tests of some optical components. Thanks to the “solar tower” facility of the Astrophysical Observatory of Arcetri (Firenze), we have been able to reproduce an environment very similar to the telescope one, testing the main LBT AO components: the pyramid sensor and the secondary deformable mirror. Thanks to this the second phase of my PhD thesis: the measure of IM applying the sinusoidal modulation technique. At first we have measured the IM using a fiber auxiliary source to calibrate the system, without any kind of disturbance injected. After that, we have tried to use this calibration technique in order to measure the IM directly “on sky”, so adding an atmospheric disturbance to the AO system. The results obtained in this PhD work measuring the IM directly in the Arcetri solar tower system are crucial for the future development: the possibility of the acquisition of IM directly on sky means that we are able to calibrate an AO system also for extremely large telescope class where classic IM measurements technique are problematic and, sometimes, impossible. Finally we have not to forget the reason why we need this: the main aim is to observe the universe. Thanks to these new big class of telescopes and only using their full capabilities, we will be able to increase our knowledge of the universe objects observed, because we will be able to resolve more detailed characteristics, discovering, analyzing and understanding the behavior of the universe components.

Ultra-low power WSNs: distributed signal processing and dynamic resource management

This thesis presents several data processing and compression techniques capable of addressing the strict requirements of wireless sensor networks. After introducing a general overview of sensor networks, the energy problem is introduced, dividing the different energy reduction approaches according to the different subsystem they try to optimize. To manage the complexity brought by these techniques, a quick overview of the most common middlewares for WSNs is given, describing in detail SPINE2, a framework for data processing in the node environment. The focus is then shifted on the in-network aggregation techniques, used to reduce data sent by the network nodes trying to prolong the network lifetime as long as possible. Among the several techniques, the most promising approach is the Compressive Sensing (CS). To investigate this technique, a practical implementation of the algorithm is compared against a simpler aggregation scheme, deriving a mixed algorithm able to successfully reduce the power consumption. The analysis moves from compression implemented on single nodes to CS for signal ensembles, trying to exploit the correlations among sensors and nodes to improve compression and reconstruction quality. The two main techniques for signal ensembles, Distributed CS (DCS) and Kronecker CS (KCS), are introduced and compared against a common set of data gathered by real deployments. The best trade-off between reconstruction quality and power consumption is then investigated. The usage of CS is also addressed when the signal of interest is sampled at a Sub-Nyquist rate, evaluating the reconstruction performance. Finally the group sparsity CS (GS-CS) is compared to another well-known technique for reconstruction of signals from an highly sub-sampled version. These two frameworks are compared again against a real data-set and an insightful analysis of the trade-off between reconstruction quality and lifetime is given.

Analysis of charge-transport properties in GST materials for next generation phase-change memory devices

The quest for universal memory is driving the rapid development of memories with superior all-round capabilities in non-volatility, high speed, high endurance and low power. The memory subsystem accounts for a significant cost and power budget of a computer system. Current DRAM-based main memory systems are starting to hit the power and cost limit. To resolve this issue the industry is improving existing technologies such as Flash and exploring new ones. Among those new technologies is the Phase Change Memory (PCM), which overcomes some of the shortcomings of the Flash such as durability and scalability. This alternative non-volatile memory technology, which uses resistance contrast in phase-change materials, offers more density relative to DRAM, and can help to increase main memory capacity of future systems while remaining within the cost and power constraints. Chalcogenide materials can suitably be exploited for manufacturing phase-change memory devices. Charge transport in amorphous chalcogenide-GST used for memory devices is modeled using two contributions: hopping of trapped electrons and motion of band electrons in extended states. Crystalline GST exhibits an almost Ohmic I(V) curve. In contrast amorphous GST shows a high resistance at low biases while, above a threshold voltage, a transition takes place from a highly resistive to a conductive state, characterized by a negative differential-resistance behavior. A clear and complete understanding of the threshold behavior of the amorphous phase is fundamental for exploiting such materials in the fabrication of innovative nonvolatile memories. The type of feedback that produces the snapback phenomenon is described as a filamentation in energy that is controlled by electron–electron interactions between trapped electrons and band electrons. The model thus derived is implemented within a state-of-the-art simulator. An analytical version of the model is also derived and is useful for discussing the snapback behavior and the scaling properties of the device.

Mars delivery service - development of the electro-mechanical systems of the Sample Fetch Rover for the Mars Sample Return Campaign

This thesis describes the development of the Sample Fetch Rover (SFR), studied for Mars Sample Return (MSR), an international campaign carried out in cooperation between the National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA). The focus of this document is the design of the electro-mechanical systems of the rover. After placing this work into the general context of robotic planetary exploration and summarising the state of the art for what concerns Mars rovers, the architecture of the Mars Sample Return Campaign is presented. A complete overview of the current SFR architecture is provided, touching upon all the main subsystems of the spacecraft. For each area, it is discussed what are the design drivers, the chosen solutions and whether they use heritage technology (in particular from the ExoMars Rover) or new developments. This research focuses on two topics of particular interest, due to their relevance for the mission and the novelty of their design: locomotion and sample acquisition, which are discussed in depth. The early SFR locomotion concepts are summarised, covering the initial trade-offs and discarded designs for higher traverse performance. Once a consolidated architecture was reached, the locomotion subsystem was developed further, defining the details of the suspension, actuators, deployment mechanisms and wheels. This technology is presented here in detail, including some key analysis and test results that support the design and demonstrate how it responds to the mission requirements. Another major electro-mechanical system developed as part of this work is the one dedicated to sample tube acquisition. The concept of operations of this machinery was defined to be robust against the unknown conditions that characterise the mission. The design process led to a highly automated robotic system which is described here in its main components: vision system, robotic arm and tube storage.