Entwicklung und Validierung der Steigbildmethode zur Differenzierung von ausgewählten Lebensmitteln aus verschiedenen Anbausystemen und Verarbeitungsprozessen

Zusammenfassung (deutsch) Seit den 1980iger Jahren wächst die Bedeutung der sog. Bildschaffenden Methoden für die Bestimmung der Qualität ökologischer Produkte. Zu diesen Methoden gehört die Biokristallisation, Steigbild und Rundfilter-Chromatographie. Die Ergebnisse dieser Methoden sind Bilder, die anhand definierter Kriterien ausgewertet werden. Bei der Biokristallisation sind es mehr oder weniger geordnete Kristallisationen auf einer Glasplatte, bei dem Steigbild zweidimensionale Strukturen auf Chromatographiepapier. In der Vergangenheit wurden die Bilder von Spezialisten ausgewertet, die nach einer längeren Schulung produktspezifische Kriterien entwickelt hatten. Im Gegensatz zur Dünnschicht-Chromatographie, wo der einzelne Stoff von der Matrix separiert wird, ist das Ziel beim Steigbild, Strukturen der möglichst ganzen Probe zu erzeugen. Die Methode wurde von Kolisko in den 1929iger Jahren entwickelt, wobei eine Kombination aus Chromatographieprozess und Metallkomplexreaktionen genutzt wurde. Die Firma WALA entwickelte die Methode für die Kontrolle ihrer Produkte und setze Silbernitrat und Eisensulfat ein. Bisher wurde die Methode qualitativ beschreibend ausgewertet, wobei einzelne Bildelemente und deren Interaktion beschrieben wurden. Deshalb musste für die vorliegende Arbeit Auswertungsmethoden entwickelt werden, mit denen auch eine statistische Bearbeitung der Ergebnisse möglich ist (nominale Unterscheidung von proben anhand der Bilder). Die Methode wurde bisher in einer Reihe von Studien eingesetzt (u.a. die Unterscheidung von Produktionsweisen). Obwohl die Bilder nur qualitativ ausgewertet wurden, konnten geschulte Prüfpersonen Proben aus verschiedenen Anbausystemen anhand der Bilder trennen. Die Ergebnisse wurden aber nicht so dokumentiert, dass sie den Erfordernissen internationaler Standardnormen für Laboratorien genügten. Deshalb mussten für diese Arbeit zunächst die Prozeduren dokumentiert und eine systematische Untersuchung zu den Einflussgrößen durchgeführt werden. Dazu wurde die visuelle Bildauswertung entwickelt und standardisiert. Die visuelle Bildauswertung basiert auf morphologischen Kriterien der Bilder von den untersuchten Weizen- und Möhrenproben. Ein Panel aus geschulten Personen entwickelte dann die Kriterien und legte sie anhand von Referenzbildern fest. Die Bilder der vorliegenden Arbeit wurden mit der einfach beschreibenden Prüfung ausgewertet, wie sie aus der sensorischen Prüfung von Lebensmitteln übernommen werden konnte. Mit geschulten und ungeschulten Prüfpersonen wurden Weizenproben und verschiedene Möhrensäfte mit der sog. Dreiecksprüfung ausgewertet (von ISO 4120). Alle Laborprozeduren wurden dokumentiert. Mit der Anwendung dieser Prozeduren wurden Vergleichsversuche mit Laboren in Dänemark und Holland (BRAD, LBI) durchgeführt. Die Ergebnisse waren sowohl für Weizen- als auch für Möhrenproben vergleichbar, wobei alle drei Labore zwischen jeweils zwei Proben unterscheiden konnten. Die systematische Untersuchung zu den Einflussgrößen zeigte, dass das Unterscheidungsvermögen der Methode vor allem von den klimatischen Bedingungen während der Steigphasen beeinflusst wird. Auch die Präkonditionierung der Papiere hat einen großen Einfluss, während die Wasserqualität (ultra-filtriert, de-ionisiert, destilliert) eine untergeordnete Bedeutung hat. Für Weizen- und Möhrenproben wurde sowohl die Wiederholbarkeit als auch die Reproduzierbarkeit getestet. Die Unterschiede in den Bildern der verschiedenen Proben waren dabei immer größer als die Variation durch Proben- und Bildwiederholung und das Labor. Die so charakterisierte Methode wurde auf kodierte Proben von definierten Feldversuchen und auf Marktproben (Paarvergleich von Anbausystemen ökologisch und konventionell) angewandt, wobei als Ergebnis mehr als 90% der Proben mit der einfach beschreibenden Prüfung anhand der Bilder unterschieden werden konnten. Die Auswertung mit der Dreiecksprüfung zeigte, dass sowohl Sorten und Verarbeitungsschritte (Saft) als auch Anbauweisen signifikant getrennt wurden. Darüber hinaus wurde die Methode auch erfolgreich auf Apfelproben angewandt. Weitere Untersuchungen müssen zeigen, ob sich das Potential der Methode, verschiedene Fragen wie die Authentizitätsprüfung von Lebensmitteln verifizieren lassen.

Observations on the Publicity and Usage of Parallel Programming Systems and Languages: A Survey Approach

In this publication, we report on an online survey that was carried out among parallel programmers. More than 250 people worldwide have submitted answers to our questions, and their responses are analyzed here. Although not statistically sound, the data we provide give useful insights about which parallel programming systems and languages are known and in actual use. For instance, the collected data indicate that for our survey group MPI and (to a lesser extent) C are the most widely used parallel programming system and language, respectively.

A User-Centric Perspective on Parallel Programming with Focus on OpenMP

The process of developing software that takes advantage of multiple processors is commonly referred to as parallel programming. For various reasons, this process is much harder than the sequential case. For decades, parallel programming has been a problem for a small niche only: engineers working on parallelizing mostly numerical applications in High Performance Computing. This has changed with the advent of multi-core processors in mainstream computer architectures. Parallel programming in our days becomes a problem for a much larger group of developers. The main objective of this thesis was to find ways to make parallel programming easier for them. Different aims were identified in order to reach the objective: research the state of the art of parallel programming today, improve the education of software developers about the topic, and provide programmers with powerful abstractions to make their work easier. To reach these aims, several key steps were taken. To start with, a survey was conducted among parallel programmers to find out about the state of the art. More than 250 people participated, yielding results about the parallel programming systems and languages in use, as well as about common problems with these systems. Furthermore, a study was conducted in university classes on parallel programming. It resulted in a list of frequently made mistakes that were analyzed and used to create a programmers' checklist to avoid them in the future. For programmers' education, an online resource was setup to collect experiences and knowledge in the field of parallel programming - called the Parawiki. Another key step in this direction was the creation of the Thinking Parallel weblog, where more than 50.000 readers to date have read essays on the topic. For the third aim (powerful abstractions), it was decided to concentrate on one parallel programming system: OpenMP. Its ease of use and high level of abstraction were the most important reasons for this decision. Two different research directions were pursued. The first one resulted in a parallel library called AthenaMP. It contains so-called generic components, derived from design patterns for parallel programming. These include functionality to enhance the locks provided by OpenMP, to perform operations on large amounts of data (data-parallel programming), and to enable the implementation of irregular algorithms using task pools. AthenaMP itself serves a triple role: the components are well-documented and can be used directly in programs, it enables developers to study the source code and learn from it, and it is possible for compiler writers to use it as a testing ground for their OpenMP compilers. The second research direction was targeted at changing the OpenMP specification to make the system more powerful. The main contributions here were a proposal to enable thread-cancellation and a proposal to avoid busy waiting. Both were implemented in a research compiler, shown to be useful in example applications, and proposed to the OpenMP Language Committee.

Computations in Relative Algebraic K-Groups

Let G be finite group and K a number field or a p-adic field with ring of integers O_K. In the first part of the manuscript we present an algorithm that computes the relative algebraic K-group K_0(O_K[G],K) as an abstract abelian group. We solve the discrete logarithm problem, both in K_0(O_K[G],K) and the locally free class group cl(O_K[G]). All algorithms have been implemented in MAGMA for the case K = \IQ. In the second part of the manuscript we prove formulae for the torsion subgroup of K_0(\IZ[G],\IQ) for large classes of dihedral and quaternion groups.

On Parallel Communicating Grammar Systems and Correctness Preserving Restarting Automata

This paper contributes to the study of Freely Rewriting Restarting Automata (FRR-automata) and Parallel Communicating Grammar Systems (PCGS), which both are useful models in computational linguistics. For PCGSs we study two complexity measures called 'generation complexity' and 'distribution complexity', and we prove that a PCGS Pi, for which the generation complexity and the distribution complexity are both bounded by constants, can be transformed into a freely rewriting restarting automaton of a very restricted form. From this characterization it follows that the language L(Pi) generated by Pi is semi-linear, that its characteristic analysis is of polynomial size, and that this analysis can be computed in polynomial time.

Selbstassemblierte Monolagen und magnetisch strukturierte Submonolagen mit dia- und paramagnetischen Molekülen sowie Einzelmolekülmagneten auf Phthalocyanin- und Subphthalocyanin-Basis

Mit dieser Arbeit wurde die Selbstassemblierung von dia- und paramagnetischen Molekülen sowie Einzelmolekülmagneten auf Goldsubstraten und magnetisch strukturierten Substraten untersucht. Dazu wurden drei verschiedene Klassen an Phthalocyaninderivaten verwendet: Diamagnetische Subphthalocyanine, paramagnetische Phthalocyaninatometalle und Diphthalocyaninatolanthanidkomplexe. Alle synthetisierten Verbindungen sind peripher thioethersubstituiert. Die Alkylketten (a: n-C8H17, b: n-C12H25) vermitteln die Löslichkeit in vielen organischen Solventien und sorgen für eine geordnete Assemblierung auf einer Oberfläche, wobei die Bindung auf Gold hauptsächlich über die Schwefelatome stattfindet. Die aus Lösung abgeschiedenen selbstassemblierten Monolagen wurden mit XPS, NEXAFS-Spektroskopie und ToF-SIMS untersucht. Bei der Selbstassemblierung auf magnetisch strukturierten Substraten stehen die Moleküle unter dem Einfluss magnetischer Streufelder und binden bevorzugt nur in bestimmten Bereichen. Die gebildeten Submonolagen wurden zusätzlich mit X-PEEM untersucht. Die erstmals dargestellten Manganphthalocyanine [MnClPc(SR)8] 1 wurden ausgehend von MnCl2 erhalten. Hier fand bei der Aufarbeitung an Luft eine Oxidation zu Mangan(III) statt; +III ist die stabilste Oxidationsstufe von Mangan in Phthalocyaninen. Der Nachweis des axialen Chloridoliganden erfolgte mit Massenspektrometrie und FIR- sowie Raman-Spektroskopie. SQUID-Messungen haben gezeigt, dass die Komplexe 1 vier ungepaarte Elektronen haben. Bei den Subphthalocyaninen [BClSubpc(SR)6] 2 wurde der axiale Chloridoligand mit dem stäbchenförmigen Phenolderivat 29-H substituiert und die erfolgreiche Ligandensubstitution durch NMR- und IR-Spektroskopie sowie Massenspektrometrie an den Produkten [BSubpc(SR)6(29)] 30 belegt. Der Radikalcharakter der synthetisierten Terbiumkomplexe [Tb{Pc(SR)8}2] 3 wurde spektroskopisch nachgewiesen; SQUID-Messungen ergaben, dass es sich um Einzelmolekülmagnete mit einer Energiebarriere U des Doppelpotentialtopfs von 880 K oder 610 cm-1 bei 3a handelt. Zunächst wurden die SAMs der Komplexverbindungen 1, 2, 30 und 3 auf nicht magnetisch strukturierten Goldsubstraten untersucht. Die Manganphthalocyanine 1 bilden geordnete SAMs mit größtenteils flach liegenden Molekülen, wie die XPS-, NEXAFS- und ToF-SIMS-Analyse zeigte. Die Mehrzahl der Thioether-Einheiten ist auf Gold koordiniert und die Alkylketten zeigen ungeordnet von der Oberfläche weg. Bei der Adsorption findet eine Reduktion zu Mangan(II) statt und der axiale Chloridoligand wird abgespalten. Das beruht auf dem sog. Oberflächen-trans-Effekt. Im vorliegenden Fall übt die Metalloberfläche einen stärkeren trans-Effekt als der axiale Ligand aus, was bisher experimentell noch nicht beobachtet wurde. Die thioethersubstituierten Subphthalocyanine 2 und 30 sowie die Diphthalocyaninatoterbium-Komplexe 3 sind ebenfalls für SAMs geeignet. Ihre Monolagen wurden mit XPS und NEXAFS-Spektroskopie untersucht, und trotz einer gewissen Unordnung in den Filmen liegen die Moleküle jeweils im Wesentlichen flach auf der Goldoberfläche. Vermutlich sind bei diesen Systemen auch die Alkylketten größtenteils parallel zur Oberfläche orientiert. Im Gegensatz zu den Manganphthalocyaninen 1 tritt bei 2b, 30a, 30b und 3b neben der koordinativen Bindung der Schwefelatome auf Gold auch eine für Thioether nicht erwartete kovalente Au–S-Bindung auf, die durch C–S-Bindungsbruch unter Abspaltung der Alkylketten ermöglicht wird. Der Anteil, zu dem dieser Prozess stattfindet, scheint nicht mit der Molekülstruktur zu korrelieren. Selbstassemblierte Submonolagen auf magnetisch strukturierten Substraten wurden mit dem diamagnetischen Subphthalocyanin 2b hergestellt. Der Nachweis der Submonolagen war schwierig und gelang schließlich durch eine Kombination von ToF-SIMS, NEXAFS Imaging und X-PEEM. Die Analyse der ToF-SIMS-Daten zeigte, dass tatsächlich eine Modulation der Verteilung der Moleküle auf einem unterwärts magnetisch strukturierten Substrat eintritt. Mit X-PEEM konnte die magnetische Struktur der ferromagnetischen Schicht des Substrats direkt der Verteilung der adsorbierten Moleküle zugeordnet werden. Die Subphthalocyanine 2b adsorbieren nicht an den Domänengrenzen, sondern vermehrt dazwischen. Auf Substraten mit abwechselnd 6.5 und 3.5 µm breiten magnetischen Domänen binden die Moleküle bevorzugt in den Bereichen geringster magnetischer Streufeldgradienten, also den größeren Domänen. Solche Substrate wurden für die ToF-SIMS- und X-PEEM-Messungen verwendet. Bei größeren magnetischen Strukturen mit ca. 400 µm breiten Domänen, wie sie aufgrund der geringeren Ortsauflösung dieser Methode für NEXAFS Imaging eingesetzt wurden, binden die Moleküle dann in allen Domänen. Die diamagnetischen Moleküle werden nach dieser Interpretation aus dem inhomogenen Magnetfeld über der Probenoberfläche heraus gedrängt und verhalten sich analog makroskopischer Diamagnete. Die eindeutige Detektion der Moleküle auf den magnetisch strukturierten Substraten konnte bisher nur für die diamagnetischen Subphthalocyanine 2b erfolgen. Um die Interpretation ihres Verhaltens bei der Selbstassemblierung in einem inhomogenen Magnetfeld weiter voranzutreiben, wurde das Subphthalocyanin 37b dargestellt, welches ein stabiles organisches TEMPO-Radikal in seinem axialen Liganden enthält. Das paramagnetische Subphthalocyanin 37b sollte auf den magnetisch strukturierten Substraten in Regionen starker magnetischer Streufelder binden und damit das entgegengesetzte Verhalten zu den diamagnetischen Subphthalocyaninen 2b zeigen. Aus Zeitgründen konnte dieser Nachweis im Rahmen dieser Arbeit noch nicht erbracht werden.

Agricultural practices and horizontal nutrient balances in urban gardens and the alternative use of urban agricultural land in Khartoum, Sudan

The surge in the urban population evident in most developing countries is a worldwide phenomenon, and often the result of drought, conflicts, poverty and the lack of education opportunities. In parallel with the growth of the cities is the growing need for food which leads to the burgeoning expansion of urban and peri-urban agriculture (UPA). In this context, urban agriculture (UA) contributes significantly to supplying local markets with both vegetable and animal produce. As an income generating activity, UA also contributes to the livelihoods of poor urban dwellers. In order to evaluate the nutrient status of urban soils in relation to garden management, this study assessed nutrient fluxes (inputs and outputs) in gardens on urban Gerif soils on the banks of the River Nile in Khartoum, the capital city of Sudan. To achieve this objective, a preliminary baseline survey was carried out to describe the structure of the existing garden systems. In cooperation with the author of another PhD thesis (Ms. Ishtiag Abdalla), alternative uses of cow dung in brick making kilns in urban Khartoum were assessed; and the socio-economic criteria of the brick kiln owners or agents, economical and plant nutritional value of animal dung and the gaseous emission related to brick making activities were assessed. A total of 40 household heads were interviewed using a semi-structured questionnaire to collect information on demographic, socio-economic and migratory characteristics of the household members, the gardening systems used and the problems encountered in urban gardening. Based on the results of this survey, gardens were divided into three groups: mixed vegetable-fodder gardens, mixed vegetable-subsistence livestock gardens and pure vegetable gardens. The results revealed that UA is the exclusive domain of men, 80% of them non-native to Khartoum. The harvested produce in all gardens was market oriented and represented the main source of income for 83% of the gardeners. Fast growing leafy vegetables such as Jew’s mallow (Corchorous olitorius L.), purslane (Portulaca oleracea L.) and rocket (Eruca sativa Mill.) were the dominant cultivated species. Most of the gardens (95%) were continuously cultivated throughout the year without any fallow period, unless they were flooded. Gardeners were not generally aware of the importance of crop diversity, which may help them overcome the strongly fluctuating market prices for their produce and thereby strengthen the contributions of UA to the overall productivity of the city. To measure nutrient fluxes, four gardens were selected and their nutrients inputs and outputs flows were monitored. In each garden, all plots were monitored for quantification of nutrient inputs and outputs. To determine soil chemical fertility parameters in each of the studied gardens, soil samples were taken from three selected plots at the beginning of the study in October 2007 (gardens L1, L2 and H1) and in April 2008 (garden H2) and at the end of the study period in March 2010. Additional soil sampling occurred in May 2009 to assess changes in the soil nutrient status after the River Nile flood of 2008 had receded. Samples of rain and irrigation water (river and well-water) were analyzed for nitrogen (N), phosphorus (P), potassium (K) and carbon (C) content to determine their nutrient inputs. Catchment traps were installed to quantify the sediment yield from the River Nile flood. To quantify the nutrient inputs of sediments, samples were analyzed for N, P, K and organic carbon (Corg) content, cation exchange capacity (CEC) and the particle size distribution. The total nutrient inputs were calculated by multiplying the sediment nutrient content by total sediment deposits on individual gardens. Nutrient output in the form of harvested yield was quantified at harvest of each crop. Plant samples from each field were dried, and analyzed for their N, P, K and Corg content. Cumulative leaching losses of mineral N and P were estimated in a single plot in garden L1 from December 1st 2008 to July 1st 2009 using 12 ion exchange resins cartridges. Nutrients were extracted and analyzed for nitrate (NO3--N), ammonium (NH4+-N) and phosphate PO4-3-P. Changes in soil nutrient balance were assessed as inputs minus outputs. The results showed that across gardens, soil N and P concentrations increased from 2007 to 2009, while particle size distribution remained unchanged. Sediment loads and their respective contents of N, P and Corg decreased significantly (P < 0.05) from the gardens of the downstream lowlands (L1 and L2) to the gardens of the upstream highlands (H1 and H2). No significant difference was found in K deposits. None of the gardens received organic fertilizers and the only mineral fertilizer applied was urea (46-0-0). This equaled 29, 30, 54, and 67% of total N inputs to gardens L1, L2, H1, and H2, respectively. Sediment deposits of the River Nile floods contributed on average 67, 94, 6 and 42% to the total N, P, K and C inputs in lowland gardens and 33, 86, 4 and 37% of total N, P, K and C inputs in highland gardens. Irrigation water and rainfall contributed substantially to K inputs representing 96, 92, 94 and 96% of total K influxes in garden L1, L2, H1 and H2, respectively. Following the same order, total annual DM yields in the gardens were 26, 18, 16 and 1.8 t ha-1. Annual leaching losses were estimated to be 0.02 kg NH4+-N ha-1 (SE = 0.004), 0.03 kg NO3--N ha-1 (SE = 0.002) and 0.005 kg PO4-3-P ha-1 (SE = 0.0007). Differences between nutrient inputs and outputs indicated negative nutrient balances for P and K and positive balances of N and C for all gardens. The negative balances in P and K call for adoptions of new agricultural techniques such as regular manure additions or mulching which may enhance the soil organic matter status. A quantification of fluxes not measured in our study such as N2-fixation, dry deposition and gaseous emissions of C and N would be necessary to comprehensively assess the sustainability of these intensive gardening systems. The second part of the survey dealt with the brick making kilns. A total of 50 brick kiln owners/or agents were interviewed from July to August 2009, using a semi-structured questionnaire. The data collected included general information such as age, family size, education, land ownership, number of kilns managed and/or owned, number of months that kilns were in operation, quantity of inputs (cow dung and fuel wood) used, prices of inputs and products across the production season. Information related to the share value of the land on which the kilns were built and annual income for urban farmers and annual returns from dung for the animal raisers was also collected. Using descriptive statistics, budget calculation and Gini coefficient, the results indicated that renting the land to brick making kilns yields a 5-fold higher return than the rent for agriculture. Gini coefficient showed that the kiln owners had a more equal income distribution compared to farmers. To estimate emission of greenhouse gases (GHGs) and losses of N, P, K, Corg and DM from cow dung when used in brick making, samples of cow dung (loose and compacted) were collected from different kilns and analyzed for their N, P, K and Corg content. The procedure modified by the Intergovernmental Panel on Climate Change (IPCC, 1994) was used to estimate the gaseous emissions of cow dung and fuel wood. The amount of deforested wood was estimated according to the default values for wood density given by Dixon et al. (1991) and the expansion ratio for branches and small trees given by Brown et al. (1989). The data showed the monetary value of added N and P from cow dung was lower than for mineral fertilizers. Annual consumption of compacted dung (381 t DM) as biomass fuel by far exceeded the consumption of fuel wood (36 t DM). Gaseous emissions from cow dung and fuel wood were dominated by CO2, CO and CH4. Considering that Gerif land in urban Khartoum supports a multifunctional land use system, efficient use of natural resources (forest, dung, land and water) will enhance the sustainability of the UA and brick making activities. Adoption of new kilns with higher energy efficiency will reduce the amount of biomass fuels (cow dung and wood) used the amount of GHGs emitted and the threat to the few remaining forests.

Pi: A Parallel Architecture Interface for Multi-Model Execution

This thesis defines Pi, a parallel architecture interface that separates model and machine issues, allowing them to be addressed independently. This provides greater flexibility for both the model and machine builder. Pi addresses a set of common parallel model requirements including low latency communication, fast task switching, low cost synchronization, efficient storage management, the ability to exploit locality, and efficient support for sequential code. Since Pi provides generic parallel operations, it can efficiently support many parallel programming models including hybrids of existing models. Pi also forms a basis of comparison for architectural components.

Parallel Methods for Synthesizing Whole-Hand Grasps from Generalized Prototypes

This report addresses the problem of acquiring objects using articulated robotic hands. Standard grasps are used to make the problem tractable, and a technique is developed for generalizing these standard grasps to increase their flexibility to variations in the problem geometry. A generalized grasp description is applied to a new problem situation using a parallel search through hand configuration space, and the result of this operation is a global overview of the space of good solutions. The techniques presented in this report have been implemented, and the results are verified using the Salisbury three-finger robotic hand.

Thread Scheduling Mechanisms for Multiple-Context Parallel Processors

Scheduling tasks to efficiently use the available processor resources is crucial to minimizing the runtime of applications on shared-memory parallel processors. One factor that contributes to poor processor utilization is the idle time caused by long latency operations, such as remote memory references or processor synchronization operations. One way of tolerating this latency is to use a processor with multiple hardware contexts that can rapidly switch to executing another thread of computation whenever a long latency operation occurs, thus increasing processor utilization by overlapping computation with communication. Although multiple contexts are effective for tolerating latency, this effectiveness can be limited by memory and network bandwidth, by cache interference effects among the multiple contexts, and by critical tasks sharing processor resources with less critical tasks. This thesis presents techniques that increase the effectiveness of multiple contexts by intelligently scheduling threads to make more efficient use of processor pipeline, bandwidth, and cache resources. This thesis proposes thread prioritization as a fundamental mechanism for directing the thread schedule on a multiple-context processor. A priority is assigned to each thread either statically or dynamically and is used by the thread scheduler to decide which threads to load in the contexts, and to decide which context to switch to on a context switch. We develop a multiple-context model that integrates both cache and network effects, and shows how thread prioritization can both maintain high processor utilization, and limit increases in critical path runtime caused by multithreading. The model also shows that in order to be effective in bandwidth limited applications, thread prioritization must be extended to prioritize memory requests. We show how simple hardware can prioritize the running of threads in the multiple contexts, and the issuing of requests to both the local memory and the network. Simulation experiments show how thread prioritization is used in a variety of applications. Thread prioritization can improve the performance of synchronization primitives by minimizing the number of processor cycles wasted in spinning and devoting more cycles to critical threads. Thread prioritization can be used in combination with other techniques to improve cache performance and minimize cache interference between different working sets in the cache. For applications that are critical path limited, thread prioritization can improve performance by allowing processor resources to be devoted preferentially to critical threads. These experimental results show that thread prioritization is a mechanism that can be used to implement a wide range of scheduling policies.

Parallel Coupled Micro-Macro Actuators

This thesis presents a new actuator system consisting of a micro-actuator and a macro-actuator coupled in parallel via a compliant transmission. The system is called the Parallel Coupled Micro-Macro Actuator, or PaCMMA. In this system, the micro-actuator is capable of high bandwidth force control due to its low mass and direct-drive connection to the output shaft. The compliant transmission of the macro-actuator reduces the impedance (stiffness) at the output shaft and increases the dynamic range of force. Performance improvement over single actuator systems was expected in force control, impedance control, force distortion and reduction of transient impact forces. A set of quantitative measures is proposed and the actuator system is evaluated against them: Force Control Bandwidth, Position Bandwidth, Dynamic Range, Impact Force, Impedance ("Backdriveability'"), Force Distortion and Force Performance Space. Several theoretical performance limits are derived from the saturation limits of the system. A control law is proposed and control system performance is compared to the theoretical limits. A prototype testbed was built using permanenent magnet motors and an experimental comparison was performed between this actuator concept and two single actuator systems. The following performance was observed: Force bandwidth of 56Hz, Torque Dynamic Range of 800:1, Peak Torque of 1040mNm, Minimum Torque of 1.3mNm. Peak Impact Force was reduced by an order of magnitude. Distortion at small amplitudes was reduced substantially. Backdriven impedance was reduced by 2-3 orders of magnitude. This actuator system shows promise for manipulator design as well as psychophysical tests of human performance.

ADAM: A Decentralized Parallel Computer Architecture Featuring Fast Thread and Data Migration and a Uniform Hardware Abstraction

The furious pace of Moore's Law is driving computer architecture into a realm where the the speed of light is the dominant factor in system latencies. The number of clock cycles to span a chip are increasing, while the number of bits that can be accessed within a clock cycle is decreasing. Hence, it is becoming more difficult to hide latency. One alternative solution is to reduce latency by migrating threads and data, but the overhead of existing implementations has previously made migration an unserviceable solution so far. I present an architecture, implementation, and mechanisms that reduces the overhead of migration to the point where migration is a viable supplement to other latency hiding mechanisms, such as multithreading. The architecture is abstract, and presents programmers with a simple, uniform fine-grained multithreaded parallel programming model with implicit memory management. In other words, the spatial nature and implementation details (such as the number of processors) of a parallel machine are entirely hidden from the programmer. Compiler writers are encouraged to devise programming languages for the machine that guide a programmer to express their ideas in terms of objects, since objects exhibit an inherent physical locality of data and code. The machine implementation can then leverage this locality to automatically distribute data and threads across the physical machine by using a set of high performance migration mechanisms. An implementation of this architecture could migrate a null thread in 66 cycles -- over a factor of 1000 improvement over previous work. Performance also scales well; the time required to move a typical thread is only 4 to 5 times that of a null thread. Data migration performance is similar, and scales linearly with data block size. Since the performance of the migration mechanism is on par with that of an L2 cache, the implementation simulated in my work has no data caches and relies instead on multithreading and the migration mechanism to hide and reduce access latencies.

Mobile P2Ping: A Super-Peer based Structured P2P System Using a Fleet of City Buses

Recently, researchers have introduced the notion of super-peers to improve signaling efficiency as well as lookup performance of peer-to-peer (P2P) systems. In a separate development, recent works on applications of mobile ad hoc networks (MANET) have seen several proposals on utilizing mobile fleets such as city buses to deploy a mobile backbone infrastructure for communication and Internet access in a metropolitan environment. This paper further explores the possibility of deploying P2P applications such as content sharing and distributed computing, over this mobile backbone infrastructure. Specifically, we study how city buses may be deployed as a mobile system of super-peers. We discuss the main motivations behind our proposal, and outline in detail the design of a super-peer based structured P2P system using a fleet of city buses.

On-the-Fly Maintenance of Series-Parallel Relationships in Fork-Join Multithreaded Programs

A key capability of data-race detectors is to determine whether one thread executes logically in parallel with another or whether the threads must operate in series. This paper provides two algorithms, one serial and one parallel, to maintain series-parallel (SP) relationships "on the fly" for fork-join multithreaded programs. The serial SP-order algorithm runs in O(1) amortized time per operation. In contrast, the previously best algorithm requires a time per operation that is proportional to Tarjan’s functional inverse of Ackermann’s function. SP-order employs an order-maintenance data structure that allows us to implement a more efficient "English-Hebrew" labeling scheme than was used in earlier race detectors, which immediately yields an improved determinacy-race detector. In particular, any fork-join program running in T₁ time on a single processor can be checked on the fly for determinacy races in O(T₁) time. Corresponding improved bounds can also be obtained for more sophisticated data-race detectors, for example, those that use locks. By combining SP-order with Feng and Leiserson’s serial SP-bags algorithm, we obtain a parallel SP-maintenance algorithm, called SP-hybrid. Suppose that a fork-join program has n threads, T₁ work, and a critical-path length of T[subscript ∞]. When executed on P processors, we prove that SP-hybrid runs in O((T₁/P + PT[subscript ∞]) lg n) expected time. To understand this bound, consider that the original program obtains linear speed-up over a 1-processor execution when P = O(T₁/T[subscript ∞]). In contrast, SP-hybrid obtains linear speed-up when P = O(√T₁/T[subscript ∞]), but the work is increased by a factor of O(lg n).

Design and Development of 3-DOF Modular Micro Parallel Kinematic Manipulator

This paper presents the research and development of a 3-legged micro Parallel Kinematic Manipulator (PKM) for positioning in micro-machining and assembly operations. The structural characteristics associated with parallel manipulators are evaluated and the PKMs with translational and rotational movements are identified. Based on these identifications, a hybrid 3-UPU (Universal Joint-Prismatic Joint-Universal Joint) parallel manipulator is designed and fabricated. The principles of the operation and modeling of this micro PKM is largely similar to a normal size Stewart Platform (SP). A modular design methodology is introduced for the construction of this micro PKM. Calibration results of this hybrid 3-UPU PKM are discussed in this paper.