Orientador: Alice Maria Bastos Hubinger Tokarnia / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-16T03:05:57Z (GMT). No. of bitstreams: 1 Tachibana_Marina_M.pdf: 2745315 bytes, checksum: 91aeb0d6708948d94d06a63e21b98ad6 (MD5) Previous issue date: 2010 / Resumo: Um sistema embarcado executa um único programa ou um conjunto pré-definido de programas repetidamente e, muitas vezes, seus componentes podem ser customizados para satisfazer uma especificação com requisitos referentes à área, desempenho e consumo de energia. Caches on-chip, em particular, são alvos de muitos algoritmos de customização por terem uma contribuição importante no desempenho e no consumo de energia de processadores embarcados. Várias aplicações embarcadas processam estruturas de dados cujos padrões de acesso distintos tornam difícil encontrar uma configuração para o cache que garanta desempenho e baixo consumo. Propomos, neste trabalho, uma metodologia para projetar caches de matrizes particionados que satisfaçam uma restrição de tamanho total e em cujas partições estão mapeadas as matrizes da aplicação. Estas partições exploram a diferença de localidade espacial entre as matrizes. Com base na simulação de rastros de acesso à memória para entradas típicas, definimos uma métrica que quantifica o uso que as matrizes fazem das metades das linhas de um cache de matrizes unificado, associativo por conjunto, que satisfaz uma restrição de tamanho. Esta métrica é usada para dividir as matrizes em dois grupos, que são mapeados em duas partições de cache, uma com mesmo tamanho de linha, e outra com metade do tamanho de linha do cache de matrizes unificado. Este procedimento é repetido para várias organizações de cache de matrizes unificados com um tamanho especificado. No final, os caches de matrizes particionados baseados em rastros de acesso à memória com menor tempo médio de acesso à memória são selecionados. Para um decodificador MPEG-2, dependendo do paralelismo dos acessos de dados, os resultados das simulações mostram que o tempo médio de acesso à memória de um cache de matrizes particionado baseado em rastros de 8K bytes apresenta uma redução de 26% a 60%, quando comparado com o cache de matrizes unificado, associativo por conjunto, de mesmo tamanho, com menor tempo médio de acesso à memória. Existe também uma redução de 46% no consumo de energia entre estes caches / Abstract: An embedded system executes a single application or a pre-defined set of applications repeatedly and, frequently, its components can be fine-tuned to satisfy a specification with requirements related to area, performance, and energy consumption. On-chip caches, in particular, are the target of several customization algorithms due to its important contribution to the performance and energy consumption of embedded processors. Several embedded applications process data structures whose access patterns turn it difficult to find a cache configuration that guarantees performance and low energy consumption. In this work, we propose a methodology for designing a split array cache that satisfies a total size constraint and in whose partitions the arrays of an application are mapped. Those partitions explore the difference in spatial locality among the matrices. Using traces of memory accesses, obtained for typical input patterns, we define a metric that quantifies the use of the two halves of the lines by array accesses in a unified array set-associative cache that satisfies a size constraint. We use this metric to split the arrays in two groups that are mapped to two cache partitions, one with the same line size, and the other with half line size of that of the unified array cache. This procedure is repeated for several unified array cache organizations of a specified size. In the end, the trace based split array caches with lowest average memory access time are selected. For a MPEG-2 decoder, depending on the parallelism of array accesses, simulation results show that the average memory access time of an 8K byte split array cache is reduced from 26% to 60% as compared to that of the unified set associative array cache of same size with the lowest average memory access time. There is also a reduction of 46% in the consumption of energy / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

In the rapidly progressing evolution of technology, more and more emphasize is put on developing proper tools for the task of designing new and revolutionary systems. These tools are required in order to allow for a designer to fully utilize the power of new architectures and techniques. This thesis examines the current state of available scheduling tools for embedded systems, by evaluating and analyzing a number of di erent tools. An attempt is made to provide an overview of how the tools are constructed, and what types of methodology have been used.

This paper is an all-embracing introduction to HTTP performance, which encompasses all areas of importance concerning application layer performance. We highlight seven areas to consider before implementing HTTP/1.1 in an embedded web server, these were identified by a careful survey of previous research. The highlighted topics are the new request methods, persistent connections, chunked encoding, byte- range operations, content negotiation, digest authentication and caching, all new or improved features in HTTP/1.1. We explore why the old design rationale of HTTP/1.1 has become obsolete in an embedded setting. We also suggest new guidelines which are more suitable for an embedded web server. / Hannes Persson, hannes.persson@mac.com Christian Toft, mda00cto@student.bth.se

In several types of embedded systems the applications are deployed both as software and as hardware components. For such systems, the partitioning decision is highly important since the implementation in software or hardware heavily influences the system properties. In the industry, it is rather common practice to take deployment decisions in an early stage of the design phase and based on a limited number of aspects. Often such decisions are taken based on hardware and software designers‟ expertise and do not account the requirements of the entire system and the project and business development constraints. This approach leads to several disadvantages such as redesign, interruption, etc. In this scenario, we see the need of approaching the partitioning process from a multiple decision perspective. As a consequence, we start by presenting an analysis of the most important and popular Multiple Criteria Decision Analysis (MCDA) methods and tools. We also identify the key requirements on the partitioning process. Subsequently, we evaluate all of the MCDA methods and tools with respect to the key partitioning requirements. By using the key partitioning requirements the methods and tools that the best suits the partitioning are selected. Finally, we propose two MCDA-based partitioning processes and validate their feasibility thorough an industrial case study.

The enormous increase in device connectivity for data and telecom devices places significant challenges on the packet processing techniques used in embedded systems such as IP stacks. Therefore, the traditional packet processing software cannot handle the line rate packet flow even for the most cutting edge devices. A solution to this problem is to allow applications to directly receive packets without passing through the normal kernel stack and drivers i.e. interface directly with the hardware. Two such open source libraries for Linux are PF_RING and Netmap. In addition Freescale has a similar technology called USDPAA. In order to satisfy the first goal of this project, a detailed analysis and evaluation of PF_RING, Netmap and USDPAA has been conducted in order to determine how they compare in relation to a number of criteria such as functionality, support, performance, ease of use, software/ hardware dependencies and project stability etc. Secondly based on the earlier work, a design is proposed that can be used to build and port an application to run on Freescale DPAA based hardware (P4080) on top of USDPAA.

Syftet med examensarbetet är att minska avdriften hos en robotgräsklippare med hjälp av en enkel GPS. En avdrift sker då robotgräsklipparen faller av från den riktning som ska följas. Vilket sker på grund av externa krafter. I rapporten är definitionen på avdrift avvikelsen från en rät linje mellan robotgräsklipparens start och stopp position. Rapporten besvarar följande frågeställningar: Vilka olika mått finns för att mäta avdrift och vilken eller vilka av dessa är att föredra i det genomförda experimentet? Hur kan robotgräsklipparens avdrift minskas med mjukvara som använder data från endast en enkel GPS? Utveckling av mjukvara har skett iterativt där varje iteration resulterar i en konfiguration som minskar avdrift. För att besvara den andra frågeställningen har experiment utförts på vardera konfiguration. Resultatet från experimentet har sedan kunnat analyseras med de mått som framkom från den första frågeställningen. Studiens resultat visar att det går att minska avdriften hos en robotgräsklippare med en enkel GPS. / The purpose of this study is to reduce the drift of a robot lawnmower using a simple GPS. A drift occurs when the robot lawnmower falls off the direction that should be followed. This happens due to external forces. In this report the definition of drift is the deviation from a straight line between the robot lawnmower’s start and stop position. The thesis answers the following research questions: What different measurements are available to measure drift and what or which of these are preferred in the experiment done in this study? How can the drift from the robot lawnmower be reduced by software that uses data from only a simple GPS? The development of software has been done iteratively where each iteration results in a configuration that reduces drift. To answer the second research question, experiments have been performed on each configuration. The result of the experiment could then be analyzed with the measurements that emerged from the first research question. The study's results show that it is possible to reduce the drift of a robotic lawnmower with a simple GPS.

Abstract Building an embedded system from an idea to a product is a slow and expensive process requiring a lot of expertise. Depending on the developer’s expertise, the required quantity and price level of the final product, and the time and money available for development, the developer can build a device from different granularity of components, ranging from ready-made platforms, kits, and modules to individual components. Generally, solutions requiring less expertise, time and money produce products with higher production costs. The main contribution of this thesis is the EOC (Embedded Object Concept) and Atomi II Framework. EOC utilizes common object-oriented methods used in software by applying them to small electronic modules, which create complete functional entities. The conceptual idea of the embedded objects is implemented with the Atomi II framework, which contains several techniques for making the EOC a commercially feasible implementation. The EOC and the Atomi II Framework decreases the difficulty level of making embedded systems by enabling a use of ready-made modules to build systems. It enables automatic conversion of a device made from such modules into an integrated PCB, lowering production costs compared to other modular approaches. Furthermore, it also enables an automatic production tester generation due to its modularity. These properties lower the number of skills required for building an embedded system and quicken the path from an idea to a commercially applicable device. A developer can also build custom modules of his own if he possesses the required expertise. The test cases demonstrate the Atomi II Framework techniques in real world applications, and demonstrate the capabilities of Atomi objects. According to our test cases and estimations, an Atomi based device becomes approximately 10% more expensive than a device built from individual components, but saves up to 50% time, making it feasible to manufacture up to 10-50k quantities with this approach.

In a modern industrial system, the requirement on computational capacity has increased dramatically, in order to support a higher number of functionalities, to process a larger amount of data or to make faster and safer run-time decisions. Instead of using a traditional single-core processor where threads can only be executed sequentially, multi-core and many-core processors are gaining more and more attentions nowadays. In a multi-core processor, software programs can be executed in parallel, which can thus boost the computational performance. Many-core processors are specialized multi-core processors with a larger number of cores which are designed to achieve a higher degree of parallel processing. An on-chip communication bus is a central intersection used for data-exchange between cores, memory and I/O in most multi-core processors. As the number of cores increases, more contention can occur on the communication bus which raises a bottleneck of the overall performance. Therefore, in order to reduce contention incurred on the communication bus, a many-core processor typically employs a Network-on-Chip (NoC) to achieve data-exchange. Real-time embedded systems have been widely utilized for decades. In addition to the correctness of functionalities, timeliness is also an important factor in such systems. Violation of specific timing requirements can result in performance degradation or even fatal problems. While executing real-time applications on many-core processors, the timeliness of a NoC, as a communication subsystem, is essential as well. Unfortunately, many real-time system designs over-provision resources to guarantee the fulfillment of timing requirements, which can lead to significant resource waste. For example, analysis of a NoC design yields that the network is already saturated (i.e. accepting more traffic can incur requirement violation), however, in reality the network actually has the capacity to admit more traffic. In this thesis, we target such resource wasting problems related to design and analysis of NoCs that are used in real-time systems. We propose a number of solutions to improve the schedulability of real-time traffic over wormhole-switched NoCs in order to further improve the resource utilization of the whole system. The solutions focus mainly on two aspects: (1) providing more accurate and efficient time analyses; (2) proposing more cost-effective scheduling methods.

Purpose – The purpose of this thesis was to facilitate the search for echoes caused by fixed objects, so called fixed echoes, when using SODAR-technique for wind measurements. Furthermore, it was investigated how fixed echoes in measurements can change based on different conditions such as: the angle of the sound beam towards the object causing the echo, the output frequency of the instrument, the air temperature and relative humidity. Method – Data was recorded on a test site in southern Sweden during February - April 2017 using two different SODAR-instruments, AQ500 and AQ510, manufactured by AQSystem. Experiments were conducted by installing the instruments at different distances and in different angles next to a met mast. The data recorded was examined for fixed echoes both by using Excel and manually and compared for correlation with the parameters mentioned. A software that uses the r-squared value for a power function adapted to the collected data was developed to detect fixed echoes. Findings – The result of the study showed that the angle of the SODAR-instrument sound beam direction compared to the met mast has a relatively high impact on the number of fixed echoes that can be easily detected. When a sound beam is aimed directly at the fixed object the echoes increased significantly compared to when the sound beams were aimed next to the mast. AQ510, that uses a higher measurement frequency than the AQ500, was less susceptible to fixed echoes when doing simultaneous measurements. Both temperature and relative humidity showed low correlation to the number of fixed echoes so the implication was that these atmospherical parameters do not affect the emergence of fixed echoes. The r-squared value for a power function adapted to the wind data turned out to be a good measure for the magnitude of a fixed echo. When the correlation coefficient of the r-squared value and the percentage of wind profiles that contained fixed echoes were calculated the value was 0,995 which is a strong positive correlation. The r-squared value is compared with a threshold value (which depends on the number of wind profiles in the data to be analysed) to determine if a fixed echo is causing disturbance. Another control is also made in the software where the measured wind values are compared with the values from the power function to find minor deviations possibly caused by fixed echoes. Implications – When the result of the thesis is considered it can help the person installing the instruments as well as the developers of the SODAR-instruments. The person installing the system can notice that fixed echoes can be decreased by rotating the instrument. The result showed the developer that by carefully selecting the right frequency the number of fixed echoes can be decreased. If neither the temperature nor the relative humidity has any impact on fixed echoes it shows the developer that the instrument can be used in varying climates. The result of the study can also be of service to wind analysts using SODAR-equipment. The software developed can be used to find fixed echoes in a more efficient way than was previously possible. Limitations – The data recordings took place during a limited period during late winter, early spring in the southern parts of Sweden which means that neither extremely high or low temperatures were measured during the campaign. The thesis is further limited by only using ”multiple axis” SODAR-instruments from AQSystem. Keywords – SODAR, wind measurements, fixed echoes, measurement instrument. / Syfte – Syftet med detta examensarbete var att underlätta sökandet efter ekon från fasta objekt, så kallade "fasta ekon", vid vindmätning med SODAR-teknik. Vidare var syftet att utreda om fasta ekon framträder olika under olika förutsättningar som: mätinstrumentets ljudlobsvinkel mot objektet som ger upphov till eko, mätfrekvens, luftens temperatur och den relativa luftfuktigheten.  Metod – Data samlades in med två olika SODAR-instrument, AQ500 och AQ510, från företaget AQSystem på en testplats i södra Sverige, februari – april 2017. Data samlades in genom att installera instrumenten på olika avstånd, samt roterat åt olika riktningar, från ett stationärt objekt (en ca 100 m hög mast). Denna data genomsöktes sedan, både manuellt och med hjälp av Excel, efter fasta ekon och jämfördes med tidigare nämnda parametrar. En mjukvara som kan användas för att detektera fasta ekon utvecklades. Mjukvaran använder r-kvadratvärdet för en potensfunktion anpassad till insamlade data för att avgöra storleken på ett eventuellt fast eko. Resultat – Studiens resultat visade att SODAR-instrumentets ljudlobsvinkel mot ekoalstrande objekt hade relativt hög påverkan på fasta ekons styrka. När varsin ljudlob på instrumenten riktades direkt mot masten ökade det fasta ekot jämfört mot när ljudloberna var riktade vid sidan av masten. AQ510, som mäter med en högre frekvens än AQ500, påverkades mindre av fasta ekon än vad AQ500 gjorde vid simultana mätningar. Både luftens temperatur och den relativa luftfuktigheten hade svag korrelation med fasta ekons styrka. Därför drogs slutsatsen att just dessa atmosfäriska parametrar ej påverkar hur fasta ekon uppstår. r-kvadratvärdet för en potensfunktion anpassad till vinddata visades vara ett bra mått på magnituden av ett fast eko. När korrelationskoefficienten för r-kvadratvärdet och andelen vindprofiler innehållande fasta ekon i vinddata beräknades antog den värdet 0,995 vilket visar på en stark positiv korrelation. r-kvadratvärdet jämförs med ett tröskelvärde (som beror av antalet vindprofiler i data som analyseras) för att avgöra om ett fast eko orsakar störningar. En kontroll görs även där det de uppmätta vindvärdena jämförs med potensfunktionens värden för att hitta mindre avvikelser som kan ha orsakats av fasta ekon. Implikationer – Om examensarbetets resultat tas i beaktning kan det underlätta för både installatörer och utvecklare av SODAR-instrument. För installatören visar resultatet att fasta ekon kan minskas genom att rotera mätinstrumentet. För utvecklaren visar resultatet att fasta ekon kan minskas genom att välja en lämplig mätfrekvens. Om varken luftens temperatur eller den relativa luftfuktigheten påverkar fasta ekon nämnvärt kan detta visa på att vindmätning med SODAR-instrument är brukbart i varierande klimat. Även analytiker kan ha nytta av examensarbetets resultat i form av den mjukvara som utvecklats. Mjukvaran kan användas för att på ett mer effektivt sätt än tidigare upptäcka fasta ekon. Begränsningar – Datainsamlingen genomfördes under en begränsad period under sen vinter till tidig vår i södra Sverige. Därför har varken exceptionellt låga- eller höga temperaturer mätts upp. Examensarbetet begränsas ytterligare av att endast mätinstrument från företaget AQSystem av typen "multiple axis" har använts. Nyckelord – SODAR, vindmätning, fasta ekon, mätinstrument. / <p>Presentationen har redan skett.</p>

Proceedings of International Summerschool Computer Science 2014