A network service provider typically sells service at a fixed traffic rate to customers. This rate is enforced by allowing or dropping packets that pass through, in a process called policing. Distributed policing is a version of the problem where a number of policers must limit their combined traffic allowance to the specified rate. The policers must coordinate their behaviour such that customers are fully allowed the rate they pay for, without receiving too much more, while maintaining some semblance of fairness between packets arriving at one policer versus another. A review of prior solutions shows that most use predictions or estimations to heuristically allocate rates, and thus cannot provide any error bounds or guarantees on the achieved rate under all scenarios. Other solutions may suffer from starvation or unfairness under certain traffic demand patterns. We present a new global ``leaky bucket'' approach that provably prevents starvation, guarantees full utilization, and provides a simple upper bound on the rate allowed under any incoming traffic pattern. We find that the algorithm guarantees a minimum 1/n share of the rate for each policer, and achieves close to max-min fairness in many, but not all cases. We also suggest some experimental modifications that could improve the fairness in practice.

A network service provider typically sells service at a fixed traffic rate to customers. This rate is enforced by allowing or dropping packets that pass through, in a process called policing. Distributed policing is a version of the problem where a number of policers must limit their combined traffic allowance to the specified rate. The policers must coordinate their behaviour such that customers are fully allowed the rate they pay for, without receiving too much more, while maintaining some semblance of fairness between packets arriving at one policer versus another. A review of prior solutions shows that most use predictions or estimations to heuristically allocate rates, and thus cannot provide any error bounds or guarantees on the achieved rate under all scenarios. Other solutions may suffer from starvation or unfairness under certain traffic demand patterns. We present a new global ``leaky bucket'' approach that provably prevents starvation, guarantees full utilization, and provides a simple upper bound on the rate allowed under any incoming traffic pattern. We find that the algorithm guarantees a minimum 1/n share of the rate for each policer, and achieves close to max-min fairness in many, but not all cases. We also suggest some experimental modifications that could improve the fairness in practice.

Thesis advisor: Dunwei Wang / Electrochemical CO2 reduction reaction (CRR) provides a solution to both the increasing global demand of energy by forming valuable chemical products for fuel production, and global warming by reducing the amount of CO2 in the environment. To efficiently reduce CO2, we sought to understand the reaction mechanism using a polycrystalline Au electrode and the super concentrated LiTFSI solution (WiS) as the electrolyte. By varying both the electrolytic potential and the concentration of WiS, we investigated the factors determining product selectivity and found that reaction kinetics and mass transport together direct the selectivity towards CO. We probed the rate limiting step (RLS) of CO2 reduction by observing the variation of product distribution with water availability in solution, and discovered that the RLS was likely to involve only a single electron transfer to form COO*–. Lastly, we proposed that in WiS, H2O were the dominant proton sources for both CO2 reduction and H2 evolution reactions. In 21m WiS, the competing hydrogen evolution reaction was kinetically inhibited, so CO production was favored with a selectivity of 90% at a potential as early as -0.4V vs RHE. This study demonstrated the great potential of WiS as a platform for studying multi-proton, multi-electron transfer reactions. / Thesis (BS) — Boston College, 2018. / Submitted to: Boston College. College of Arts and Sciences. / Discipline: Scholar of the College. / Discipline: Chemistry.

No description available.

<p> In this thesis, the rate-limiting step is determined in a biofuel cell with a bio-anode, a Nation membrane and a conventional, platinum-based cathode using reference electrode method. It was discovered by surprise that the cathode overpotential dominated the cell overpotential. Na + in the membrane was found to hinder the W transport. The cathode overpotential increased due to the presence of Na + in the membrane and at the cathode. The limited H+ transport causes the increase of the cathode overpotential. H+ transport is the rate-limiting step in our biofuel cell, rather than commonly believed electron transport. Moreover, the cell power output degradation is not due to the conventionally believed depletion of the fuel substrate, inter-penetration of the fuel and oxidizer and the degradation of the biocatalysts, but the limited W transport in our biofuel cell. </p> <p> The existing oxygen reduction mechanism at the cathode was questioned and revised. When Na+ occupies all sulfonate groups in the membrane, only the Na+ from the buffer can pass through the membrane. The oxygen reacts with the water transported with Na+ and electrons to produce OH", which balances with the transported Na+ to keep electroneutrality at the cathode. </p> <p> Tris buffer without Na + was utilized as alternative anolyte in the biofuel cell. It was found that the cell with Tris buffer had a poorer performance in comparison with sodium phosphate buffer due to the increases of the anode and cathode overpotentials. Tris buffer does not constitute a solution to the problem. </p> <p> This work represents a step toward a more complete understanding of the properties of biofuel cells. To improve biofuel cell output, the herein identified H+ transport limitation in Na + contained Nation needs to be overcome. </p> / Thesis / Master of Applied Science (MASc)

<p> Intercritically annealed steels, such as dual-phase steels, have found widespread use in automotive structural components due to their high strength and ductility. Elements such as Mn, Al and Si, added to improve the mechanical properties are selectively oxidized during heat treatment and limit the ability of the alloy to be reactively wet during continuous hot-dip galvanizing. Subsequently, a limit has been placed on the amount of alloy which can be used if the steel is to be subsequently galvanized. The specifics of this limit have not been explored in detail, nor has the mechanism of decreased wettability been well demonstrated in the literature other than to say that the galvanizing reaction is limited by oxides on the surface.</p> <p> Using a force balance, it is shown that the presence of MnO on the surface of steels greatly reduces the wettability with a typical galvanizing bath (Zn-0.2wt%Al, Fe-saturated, 460°C). Furthermore, it was determined that this is caused by the additional and rate-limiting step of aluminothermic reduction of the oxide layer with the bath Al, required for subsequent inhibition layer formation. By using a low pO2 during annealing, the wettability was improved by reducing the thickness of the MnO layer when compared to intermediate and industrially common values of pO2. Using a high pO2 also resulted in improved wettability since the internal oxide which was formed did not reduce the wettability since it was not exposed to the bath alloy.</p> <p> Improvements in hardenability were also explored via dilatometry showing that the formation of bainite is delayed with increasing Mn content, as well as a decrease in transformation temperatures from γ during cooling (i.e. Ms and Bs). At ~5wt% Mn, only the the transformation to αM could be observed. This opens the door to higher strength, galvanized steels - as well as possibly galvanized martensitic steels.</p> / Thesis / Master of Applied Science (MASc)

ABSTRACT RATE-LIMITING STEP OF CONE PHOTOTRANSDUCTION RECOVERY AND OGUCHI DISEASE MECHANISMS By Frank Sungping Chen Advisor: Ching-Kang Jason Chen, Ph.D. Retinal photoreceptors provide the first gateway in which light information from the environment is transformed into neuronal signals. The cone and rod photoreceptors are responsible for day and night vision, respectively. Understanding rod and cone phototransduction is to figure out how these cells differ in their temporal and spatial sensitivities to allow perception of a broad dynamic range of stimuli. Phototransduction is mediated through a Gprotein signaling cascade. Light absorption by visual pigment triggers the isomerization of 11- cis-retinal covalently attached to these pigments, which are heptahelical transmembrane Gprotein- coupled receptors. Isomerization of 11-cis-retinal to all-trans-retinal activates the receptor, which catalyzes the exchange of GDP for GTP on the α subunit of heterotrimeric Gprotein called transducin. Activated transducin relieves inhibitory constraint on cGMP-PDE, leading to rapid hydrolysis of cGMP, closure of cGMP gated cation channels, and membrane hyperpolarization. In order for photoreceptor to be responsive to light again, this robust phototransduction pathway must be deactivated in a timely fashion and this involves several reactions simultaneously. First, the activated opsin must be phosphorylated by G-protein-coupled receptor kinases (GRKs) and capped by arrestin binding. Second, activated transducin must hydrolyze bound GTP through intrinsic GTPase activity, which is accelerated by a GTPase accelerating protein (GAP) complex comprised of RGS9-1/Gβ5-L/R9AP. Mutations in human genes involved in these reactions cause various visual defects. Cone, by and large, uses the same set of genes for pigment and transducin deactivations but it has lower sensitivity and faster kinetics than rod and is responsible for high visual acuity. During phototransduction recovery in which multiple reactions take place, the slowest reaction will determine the overall rate of recovery. In rod, this so-called, rate-limiting step has been determined to be transducin deactivation. It is unknown whether cone transducin deactivation also controls the timing of conerecovery, although we and others have shown that cone possesses a higher level of GAP concentration. In this thesis, the rate-limiting step in cone phototransduction recovery has been unequivocally determined by overexpressing RGS9-1 by 2.7 fold in mouse cones, which results in accelerated cone recovery. Complementarily, we find that ectopically expressing a human cone opsin kinase GRK7 in mouse cones does not affect cone recovery. These results altogether demonstrate that the rate-limiting step of cone recovery is the GTP-hydrolysis of cone transducin, not the opsin phosphorylation by GRKs. By elucidating the rate-limiting step of photoreceptor recovery, we have revealed the importance of G-protein cycling in timing of both rod and cone photoreceptors. This may further be generalized to other physiological processes controlled by heterotrimeric G-proteins. The proper shutoff of phototransduction is essential for normal vision as recovery defects lead to visual impairment. Even though the reaction catalyzed by GRK1 is not rate-limiting, mutations of this important gene render rhodopsin phosphorylation and deactivation the slowest step in rod recovery and create a pathological condition. GRK1 mutations have been found in Oguchi disease patients, who suffer from congenital stationary night blindness. One of the mutations, V380D, is investigated in detail in this study. Transgenic expression of GRK1 V380D mutant in rods reveals a kinase with reduced expression and catalytic activity. While V380D GRK1 is found capable of inactivating rhodopsin, the reduction in kinase activity leads to a delayed dark adaptation, and is congruent with the night blindness phenotype observed in Oguchi disease patients. Finally, we have also investigated the role of post-translational isoprenylation on GRK1 function. We found that isoprenylation is required for GRK1 membrane association and outer segment targeting. Altogether our data add significantly to understanding the structure and function of GRK1, which is one of the least understood molecules involved in vertebrate phototransduction.

Large scale internet services are increasingly implemented as distributed systems to achieve availability, fault tolerance and scalability. To achieve fault tolerance, microservices need to employ different resilience mechanisms such as automatic retries, rate limiting, circuit breaking amongst others to make the services handle failures gracefully and cause minimum damage to the performance of the overall system. These features are provided by service proxies such as Envoy which is deployed as a sidecar (sidecar proxy is an application design pattern which abstracts certain features, such as inter-service communications, monitoring and security, away from the main architecture to ease the tracking and maintenance of the application as a whole), the service proxies are very new developments and are constantly evolving. Evaluating their presence in a system to see if they add latency to the system and understand the advantages provided is crucial in determining their fit in a large scale system.Using an experimental approach, the services are load tested with and without the sidecar proxies in different configurations to determine if the usage of Envoy added latency and if its advantages overshadow its disadvantages. The Envoy sidecar proxy adds latency to the system; however, the benefits it brings in terms of resilience, make the services perform better when there is a high number of failures in the system. / Storskaliga internettjänster implementeras alltmer som distribuerade system för att uppnå tillgänglighet, feltolerans och skalbarhet. För att uppnå feltolerans måste microservices använda olika typer av resiliens mekanismer som automatisk återförsök, hastighetsbegränsning, kretsbrytning bland annat som tillåter tjänsterna att hantera misslyckanden graciöst och orsaka minimala skador på prestandan hos det övergripande systemet. Dessa funktioner tillhandahålls av service proxies som Envoy. Dessa proxies används som sidovagn (sidvagnproxy är ett applikationsdesignmönster som abstraherar vissa funktioner, såsom kommunikation mellan kommunikationstjänster, övervakning och säkerhet, bort från huvudarkitekturen för att underlätta spårningen och underhåll av ansökan som helhet). Dessa tjänster är väldigt nya och utvecklas ständigt. Att utvärdera deras närvaro i ett system för att se om de lägger till latens för systemet och förstå fördelarna som tillhandahålls är avgörande för att bestämma hur väl de skulle passa i ett storskaligt system. Med hjälp av ett experimentellt tillvägagångssätt testas tjänsterna med och utan sidospårproxys i olika konfigurationer för att avgöra om användningen av Envoy lägger till latens och om dess fördelar överskuggar dess nackdelar. Envoy sidecar proxy ökar latensen i systemet; De fördelar som det ger med avseende på resiliens gör tjänsterna bättre när det finns ett stort antal misslyckanden i systemet.

High temperature proton conducting perovskite oxides are very attractive materials for applications in electrochemical devices, such as solid oxide fuel cells (SOFCs) and hydrogen permeation membranes. A better understanding of the hydrogen oxidation mechanism over the metal/proton conductor interface, is critical for rational design to further enhance the performances of the applications. However, kinetic studies focused on the metal/proton system are limited, compared with the intensively studied metal/oxygen ion conductor system, e.g., Ni/YSZ (yttrium stabilized zirconia, Zr₁-ₓYₓO₂-δ). This work presents an elementary kinetic model developed to assess reaction pathway of hydrogen oxidation/reduction on metal/proton conductor interface. Individual rate expressions and overall hydrogen partial pressure dependencies of current density and polarization resistance were derived in different rate limiting cases. The model is testified by tailored experiments on Pt/BaZr₀.₁Ce₀.₇Y₀.₁Yb₀.₁O₃-δ (BZCYYb) interface using pattern electrodes. Comparison of electrochemical testing and the theoretical predictions indicates the dissociation of hydrogen is the rate-limiting step (RLS), instead of charge transfer, displaying behavior different from metal/oxygen ion conductor interfaces. The kinetic model presented in this thesis is validated by high quantitative agreement with experiments under various conditions. The discovery not only contributes to the fundamental understanding of the hydrogen oxidation kinetics over metal/proton conductors, but provides insights for rational design of hydrogen oxidation catalysts in a variety of electrochemical systems.

Nucleoside 5'-Diphosphate Kinase (NdPK EC 2.7.4.6) is an enzyme (phosphotransferase) with extraordinary characteristics due to its unique ability to transfer phosphor groups to interconvert all nucleoside di- and triphosphates as a part of the DNA synthesis. Due to Biovica International AB's use of signaling and complementing enzymes in their in vitro diagnostic (IVD) test assays for Thymidine Kinase activity, an investigation was proposed to evaluate NdPK, which is a complementing enzyme in the assay. The aim of the study was to evaluate the enzymatic turnover of the enzyme NdPK with a spectrophotometric assay to obtain the specific activity (Units/mg solid protein). To determine the specific activity, enzyme kinetic methodology was applied, including the Michaelis-Menten model. In this study, the method is proposed as a general internal control procedure for the company, as a tool for function control of the different purchased enzymes used in their products in development. Results from the study reflects the different methods used to gain the specific activity for NdPK, where they were compared with the already specified specific activity from the manufacturing company. The results were auspicious, but before the method's authorization as an internal quality procedure, a few amendments are in mind. For instance, determining a method for the graphical readings, validating the method for quality control, and investigating if the method is applicable to other complementing enzymes. In conclusion, the method for determining the specific activity of the enzyme NdPK can be done, by executing the procedure of colorimetric enzyme assay.