Search

The technological evolution of the fifth generation (5G) and beyond wireless networks not only enables the ubiquitous connectivity of massive user equipments (UEs), i.e., smartphones, laptops, tablets, but also boosts the development of various kinds of emerging applications, such as smart navigation, augmented reality (AR), virtual reality (VR) and online gaming. However, due to the limited battery capacity and computational capability such as central processing unit (CPU), storage, memory of UEs, running these computationally intensive applications is challenging for UEs in terms of latency and energy consumption. In order to realize the metrics of 5G, such as higher data rate and reliability, lower latency, energy reduction, etc, mobile edge computing (MEC) and unmanned aerial vehicles (UAVs) are developed as the key technologies of 5G. Essentially, the combination of MEC and UAV is becoming more and more important in current communication systems. Precisely, as the MEC server is deployed at the edge network, more and more applications can benefit from task offloading, which could save more energy and reduce round trip latency. Additionally, the implementation of UAV in 5G and beyond networks could play various roles, such as relaying, data collection, delivery, SWIFT, which can flexibly enhance the QoS of customers and reduce the load of network. In this regard, the main objective of this thesis is to investigate the UAV-enabled MEC system, and propose novel artificial intelligence (AI)-based algorithms for optimizing some challenging variables like the computation resource, the offloading strategy (user association) and UAVs' trajectory. To achieve this, some of existing research challenges in UAV-enabled MEC can be tackled by some proposed AI or DRL based approaches in this thesis. First of all, a multi-UAV enabled MEC (UAVE) is studied, where several UAVs are deployed as flying MEC platform to provide computing resource to ground UEs. In this context, the user association between multiple UEs and UAVs, the resource allocation from UAVs to UEs are optimized by the proposed reinforcement learning-based user association and resource allocation (RLAA) algorithm, which is based on the well known Q-learning method and aims at minimizing the overall energy consumption of UEs. Note that in the architecture of Q-learning, a Q-table is implemented to restore the information of all state and action pairs, which will be kept updating until the convergence is obtained. The proposed RLAA algorithm is shown to achieve the optimal performance with comparison to the exhaustive search in small scale and have considerable performance gain over typical algorithms in large-scale cases. Then, in order to tackle the more complicated problems in UAV-enabled MEC system, we first propose a convex optimization based trajectory control algorithm (CAT), which jointly optimizes the user association, resource allocation and trajectory of UAVs in the iterative way, aiming at minimizing the overall energy consumption of UEs. Considering the dynamics of communication environment, we further propose a deep reinforcement learning based trajectory control algorithm (RAT), which deploys deep neural network (DNN) and reinforcement learning (RL) techniques. Precisely, we apply DNN to optimize the UAV trajectory with continuous manner and optimize the user association and resource allocation based on matching algorithm. It performs more stable during the training procedure. The simulation results prove that the proposed CAT and RAT algorithms both achieve considerable performance and outperform other traditional benckmarks. Next, another metric named geographical fairness in UAV enabled MEC system is considered. In order to make the DRL based approaches more practical and easy to be implemented in real world, we further consider the multi agent reinforcement learning system. To this end, a multi-agent deep reinforcement learning based trajectory control algorithm (MAT) is proposed to optimize the UAV trajectory, in which each of UAV is instructed by its dedicated agent. The experimental results prove that it has considerable performance benefits over other traditional algorithms and can flexibly adjusts according to the change of environment. Finally, the integration of UAV in emergence situation is studied, where an UAV is deployed to support ground UEs for emergence communications. A deep Q network (DQN) based algorithm is proposed to optimize the UAV trajectory, the power control of each UE, while considering the number of UEs served, the fairness, and the overall uplink data rate. The numerical simulations demonstrate that the proposed DQN based algorithm outperforms the existing benchmark algorithms.

Two-component systems (TCS) are the predominant signal transduction pathways in prokaryotes, being present also in eukaryotic organisms, such as algae, fungi and yeast, and higher plants. TCSs play an important role in environmental signal perception and response, essentially implementing adaptation to the surrounding environment. Histidine Kinase 2 (Hik2) in cyanobacteria is a typical sensor histidine kinase, one component of a TCS, and has been identified to be a homologue protein of Arabidopsis Chloroplast Sensor Kinase (CSK). Previous research has elucidated Hik2 to regulate photosynthetic gene transcription with two response regulators, Rre1 and RppA via phosphorylation. A typical histidine kinase contains a variable sensor domain and a conserved kinase domain. It usually functions as a homodimer. This thesis describes the structural characterisation of Hik2, probing particularly its discovered oligomeric states. Results obtained from size exclusion chromatography, native-PAGE, chemical cross-linking analyses and mass spectrometry, amongst others, have shown a variety of Hik2 structural populations exist, further validated by negative stain transmission electron microscopy coupled to single particle analysis. Hik2 protein exists predominantly as a hexamer in low salt conditions, and adding NaCl dissociates hexamers into tetramers, critical for the autophosphorylation activity of Hik2. Thus, a model is proposed for the constitution change of Hik2 oligomers when salt concentration differs. In addition, the sensor domain is typically responsible for detecting environmental input, however, it is not yet clear how Hik2 and CSK sense signals. In this thesis, the structures of Hik2 and CSK sensor domains were analysed and discussed, to aid our understanding of their mechanism of signal perception and transduction.

Proteins are large biological molecules that perform a vast array of functions within living organisms. Proteins are made from a process called translation, in which a ribosome decodes mRNA, a single-stranded copy of DNA, to produce a specific amino acid chain. Given the essential role of proteins in maintaining life, it is of central importance to comprehend the translation process and how it is regulated. Translation process can be divided into three major stages: initiation, elongation and termination. Regulation can occur at any of these stages to control protein production. In most cases, regulation primarily targets the initiation stage. Another direct mechanism to ensure accurate protein level in the cell is to regulate the stability of mRNA. Using mathematical modelling, in this thesis we investigate how protein production is controlled. We use a stochastic modelling approach called the Totally Asymmetric Simple Exclusion Process to model the translation process. Numerical simulation acts as a complementary tool. We first investigate how mRNA stability affects protein production from one mRNA during its lifetime. Next, we investigate auto-negative feedback on translation initiation and its substantial impact on controlling protein level in a cell. Finally, we incorporate ribosome recycling into the auto-negative feedback control. Novel results such as oscillation and bistable switching in protein level are found via this mathematical analysis. These predictions invite experimental testing.

In the information age, the advent of Internet technologies has made it possible to transform language learning and teaching, through online intercultural exploration and exchange. The role of English as an international language in global communication is recognised in China's latest national guidelines for tertiary language curricula, which seek to develop professionals possessing both sound language proficiency and intercultural competence, alongside other goals such as information literacy and learner autonomy. This thesis examines the actual delivery of an intercultural dimension mediated by Internet technologies. To this end, this thesis reports an investigation into the situation of Internet-mediated intercultural English language teaching and learning at China's higher education institutions conducted between 2008 and 2009. Multi-stage and multi-site fieldwork combined a survey approach and a collective case study approach. The researcher being the primary data collection tool, data obtained from questionnaire survey, interview (and informal conversation), observation and document collection instruments have been combined. The survey findings thus not only have breadth, but also provide materials for in-depth studies of sample cases. Four cases, demonstrating a variety of institutional, individual, pedagogical and technological factors, are analysed at greater length to explore whether evidenced Internet-mediated practices might facilitate an intercultural approach to teaching and learning. Comparisons suggest that, despite their differing characteristics and contexts, there are some commonalities in terms of Internet-mediated activities entailing elements of an intercultural approach. These elements are synthesised and mapped out in an original pedagogical framework for Internet-mediated intercultural teaching and learning, with a set of guiding principles. This framework advocates establishing an Internet-mediated intra-class community as the basis for undertaking intercultural language activities, complementing the prevailing telecollaborative model for the development of intercultural communicative competence. This outcome contributes to a fuller understanding of the design of Internet-mediated intercultural language activities.

Urban areas are usually hotter than their surrounding rural areas, which is the well-known urban heat island (UHI) effect. With continuing urbanization, more people are threatened by the high thermal risks in urban areas, especially during heat waves (HWs). While there has been progress in improving our understanding of the physical processes controlling the urban thermal conditions, key questions remain, especially regarding the relative importance of different physical processes and their spatial variability. To bridge the gap, this dissertation aims to determine the key physical mechanisms that cause surface and near-surface air UHIs, and to quantify the contributions of different physical processes to the daytime warming rate of the urban boundary layer (UBL). In doing so, surface and boundary layer heat budget analyses are conducted using both observations and simulations with the Weather Research and Forecasting (WRF) model. In the first chapter of this dissertation, we quantify the magnitude of surface urban heat islands (SUHIs), or sometimes surface urban cool islands (SUCIs), and elucidate their biophysical and atmospheric drivers based on observational data collected from one urban site and two rural grassland sites in and near the city of Nanjing, China. The study highlights that the magnitude of SUHIs and SUCIs can vary strongly with the biophysical characteristics of the rural land. In the second chapter, we study the WRF-simulated surface and near-surface air UHIs during heat waves in two cities with contrasting climates (Boston and Phoenix). We demonstrate that the magnitude of UHIs or UCIs during HWs is strongly controlled by urban-rural differences in terms of aerodynamic features, moisture availability and heat storage, which show contrasting characteristics in different regions. In the last chapter, we further apply the UBL heat budget analysis to the WRF simulated results and show that in both Boston and Phoenix the surface sensible heat flux dominates the daytime UBL warming rate, whose behavior is the key to understanding how the daytime UBL warming rate changes throughout HWs. In summary, this dissertation improves our understanding of the physical processes modulating the thermal conditions of urban areas from the ground to the top of the boundary layer and provides scientific guidance on model development and mitigation of extreme heat conditions in cities.

Epstein Barr-virus (EBV) is a ubiquitous herpesvirus that causes infectious mononucleosis and is associated with a variety of human cancers. Ordinarily, EBV persists in healthy individuals by establishing latent infection in the B-cell compartment. Immunosuppression can result in unchecked outgrowth of latently infected cells, culminating in lymphoproliferative disorders and lymphomas. How EBV induces and supports the physiological shift from quiescence to long-term growth and proliferation as modeled by in vitro lymphoblastic transformation is unknown. Multiplexed proteomics on de novo infected primary B-cells was performed to uncover metabolic and signaling pathways that facilitate outgrowth and survival of B-cells latently infected with EBV. By combining chemical inhibitor studies and clustered regularly interspaced short palindromic repeats (CRISPR) genetic perturbation experiments, we discovered viral upregulation of mitochondrial one-carbon (1C) metabolism and de novo synthesis of serine, cholesterol and fatty acids, to be crucial for EBV-positive B-cell proliferation and viability. We demonstrated that mitochondrial 1C metabolism broke down serine to generate formate for nucleotide synthesis, glycine for glutathione production and cellular reducing power in the form of NADPH. We also showed that intramitochondrial NADPH was important for efficient LCL proliferation. De novo serine synthesis (DNSS) may augment 1C metabolism and/or buffer the infected cells from external fluctuations in serine availability. Both mitochondrial 1C metabolism and DNSS depended on EBV-encoded nuclear antigen 2 (EBNA2) and its target MYC for activation. EBV-infected cells also utilized cholesterol and fatty acid synthetic pathways to produce geranylgeranyl pyrophosphate (GGPP) and palmitate, respectively, to reorganize the signaling milieu of the transforming cell. Rab geranylgeranyltransferase was specifically required for EBV-positive B-cell outgrowth and proliferation. We also identified RAB13 as a potent driver of latently infected cell growth and survival, whose expression was driven by EBNA3C and likely functioned as a chaperone for viral latent membrane proteins (LMPs) to maintain the latter’s oncogenic signaling capacities. Taken altogether, our results provide an expansive investigation into the pathways that underlie the transformative change of a B lymphocyte from quiescence to activation and growth and identify vulnerabilities in EBV-driven lymphoblastic disorders that may represent new therapeutic targets.

This thesis is concerned with transformation-based statistical methods in different three areas. For problem of nonparametric regression, the transformed-based technique is called Model-free bootstrap method; for density nonparametric estimation we propose transformed flat-top series estimator; for Bayesian hypothesis testing, we employ a parametric transformation to help in evaluating the system upgrade of glucose monitoring device. First, we establish the theory of the asymptotic validity of Model-free method on construction of confidence interval under specified assumptions. The spirit of Model-free method is transforming non-i.i.d. data to i.i.d data and then bootstrapping the new i.i.d. data. We also conduct simulations to check the finite sample properties of Model-free estimators, compared to regular normal approximation and local bootstrap method. Next, we address the problem of nonparametric estimation of a smooth univariate density on compact support. If the density function has compact support and is non-zero at either boundary, regular kernel estimator will be seriously biased. A lot of bias correction methods were proposed to improve the bias on the boundary. In this chapter we propose the transformed flat-top series estimator, which keeps the same bias order as existing methods at boundary, and improves the bias in the interior region of the support to higher order. Theoretical analysis and simulations are provided, and the results are generally better than corresponding results of many other kernel density estimator with boundary correction. At last we propose modified Bayesian hypothesis method that can be assessed by type I&II errors and an alternative assessment of standard Bayesian hypothesis testing method, particularly suited to situations where modifications are made to continuous glucose monitoring (CGM) systems already approved. A parametric transformation of observations is the key to test the validation of this system upgrade. Simulations are conducted to assess the risks and benefits of the approach, which show that by careful planning and analysis prospective study sizes can be reduced and better decisions can be made on the effectiveness and safety of the modified systems

This dissertation addresses the problem of inferring scene depth information from a collection of calibrated images taken from different viewpoints via stereo matching. Although it has been heavily investigated for decades, depth from stereo remains a long-standing challenge and popular research topic for several reasons. First of all, in order to be of practical use for many real-time applications such as autonomous driving, accurate depth estimation in real-time is of great importance and one of the core challenges in stereo. Second, for applications such as 3D reconstruction and view synthesis, high-quality depth estimation is crucial to achieve photo realistic results. However, due to the matching ambiguities, accurate dense depth estimates are difficult to achieve. Last but not least, most stereo algorithms rely on identification of corresponding points among images and only work effectively when scenes are Lambertian. For non-Lambertian surfaces, the "brightness constancy" assumption is no longer valid. This dissertation contributes three novel stereo algorithms that are motivated by the specific requirements and limitations imposed by different applications. In addressing high speed depth estimation from images, we present a stereo algorithm that achieves high quality results while maintaining real-time performance. We introduce an adaptive aggregation step in a dynamic-programming framework. Matching costs are aggregated in the vertical direction using a computationally expensive weighting scheme based on color and distance proximity. We utilize the vector processing capability and parallelism in commodity graphics hardware to speed up this process over two orders of magnitude. In addressing high accuracy depth estimation, we present a stereo model that makes use of constraints from points with known depths - the Ground Control Points (GCPs) as referred to in stereo literature. Our formulation explicitly models the influences of GCPs in a Markov Random Field. A novel regularization prior is naturally integrated into a global inference framework in a principled way using the Bayes rule. Our probabilistic framework allows GCPs to be obtained from various modalities and provides a natural way to integrate information from various sensors. In addressing non-Lambertian reflectance, we introduce a new invariant for stereo correspondence which allows completely arbitrary scene reflectance (bidirectional reflectance distribution functions - BRDFs). This invariant can be used to formulate a rank constraint on stereo matching when the scene is observed by several lighting configurations in which only the lighting intensity varies.

Reverse-flow combustors have been used in heavy land-based gas turbines for many decades. A sheath is typically installed to provide cooling at an expense of large pressure losses, through small jet impingement cooling and strong forced convention channel flow. With the modern advancement in metallurgy and thermal-barrier coating technologies, it may become possible to remove this sheath to recover the pressure losses without melting the combustor chamber. However, without the sheath, the flow inside the dump diffuser may exert nonuniform cooling on the combustion chamber. Therefore, the objective of this project is to investigate the flow pattern, pressure drop, and heat transfer in the dump-diffuser reverse-flow combustor with and without sheath to determine if the sheath could be removed. The investigation was conducted through both experimental and computational simulation. The results show that 3.3% pressure losses could be recovered and the highest wall temperature will increase 18% without the sheath.

This dissertation is comprised of three autonomous chapters on topics in labor economics. The first chapter exploits the quasi-random assignment of students into classrooms in a large secondary school in Malaysia to estimate the effects of peers on student outcomes. The estimates show that having better achieving classmates improves a student's math achievement and reduces the student's incidence of class absences and discipline violations. There is also evidence of non-linear peer effects and that average achievement may increase as a result of ability grouping. The second chapter extends Iannaccone's (1992) religious club model to explain why the Amish would collectively object to high school education and refuse to comply with compulsory schooling laws. I utilize the surprising 1972 U.S. Supreme Court's decision in Wisconsin vs. Yoder, which exempts Amish children from compulsory high school education, as a policy shock to test the predictions of the model. The results show that the successful restriction on high school education helped the Amish sect exclude individuals who have high labor productivity and would lower the quality of the sect from joining. The evidence supports the idea that the Amish use the restriction on secular education as a religious sacrifice to screen out uncommitted members. The third chapter investigates the effect of higher immigration on native fertility. Previous research shows that immigration affects wages, income, and the cost of child rearing, while standard fertility model predicts that changes in wages, income, and the cost of child rearing would affect fertility. Using the cross- state variation in the total fertility rates of native- born American women and the share of immigrants in the population between 1970 and 2005, this chapter estimates that for every one percentage point increase in the share of immigrants in the population, native total fertility rate is predicted to increase by roughly 0.01 children. The negative effect of immigration on wages is the most likely explanation, because the fertility of less educated women and women who resided in their states of birth is most affected

Automatic spoken language identification (UD) refers to the task of identifying the language spoken in a particular utterance by an unknown speaker. This thesis presents several novel techniques to improve the robustness of the language identification system, especially while handling a large number of target languages. The parallel phoneme recognition followed by language modeling (PPRLM) architecture represents one of the most successful approaches to the LID task, hence the PPRLM is used as the baseline LID system in this thesis. Two schemes are proposed to improve the performance of the baseline system, namely the phoneme inventories modification and the discounting method for language modeling. A novel prosodic feature based tonal and non-tonal language classification system is described. The tonal and non-tonal language classification system can also serve as the pre-classification for the PPRLM UD system, to maintain the robustness of the LID system while handling a large number of target languages. A novel use of the Kohonen self-organizing feature map (KSFM) in the LID task is also proposed. The performance of KSFM, multi-layer KSFM (MLKSFM) and the hierarchical MLKSFM are examined. With a novel segment-based, normalized feature vector and a properly selected competitive layer topology. the multi-layer KSFM shows a promising LID accuracy of82.7% compared to the baseline system of79.2% on an 11language task. Due to the instability of the KSFM based system when the classification is performed with a large number of classes, the MLKSFM is employed as the pre-classification for the LID task. For the PPRLM with MLKSFM pre-classification, a better LID accuracy of 75.4% is obtained compared to the PPRLM with the tonal and non-tonal language pre-classification of72.5% on a 16-language task. In further experiments for improving the robustness of the LID system, the phase spectrum features - the modified group delay features (MODGDF), and the Hilbert-Huang Transform (HHT based features are examined. The parallel broad phone class recognition followed by language modeling (PBPRLM) is investigated in order to remove the bias caused by the inequality for the sizes of the phoneme inventories across different languages. The Gaussian Mixture Model based fusion technique is proposed to obtain a more robust likelihood score. With the use of all the novel techniques proposed in this thesis, an Equal Error Rate (EER) of 6.6% on the 2005 NIST LRE and a LID identification rate of 84.3% on a 22-1anguage task can be achieved.

Scientists have identified agricultural fertilizers as a primary culprit behind the eutrophication phenomenon booming in lakes and gulfs, among which the wastewater flushed from confined swine production and applied to cropland as fertilizers is identified to be one of the major responsible agricultural sources. When the scale of swine production keeps rising, manure that cannot be land applied according to the plant and soil testing has to be treated before discharge. Shortcut nitrification and denitrification (SND) is a novel nitrogen removal process that has drawn significant attention from researchers lately. In this study, the application of SND in swine wastewater treatment was investigated using two sequencing batch reactors (SBRs) connected in series in order to reduce the needs for energy and carbon uses. The first SBR produced a nitrite rich effluent that was subsequently fed to the second SBR where both nitrogen and phosphorus removal were taking place. Shortcut nitrification (also termed nitritation) is the first step of shortcut nitrogen removal from swine wastewater. As such, stably obtaining an effluent with a significant amount of nitrite becomes the premise to realize SND. The possibility of accumulating nitrite from swine wastewater was firstly investigated by adopting a continuous feeding strategy in an activated sludge SBR. The results showed that free ammonia and free nitrous acid in the system could reduce the activities of nitrite oxidizing bacteria, generating an effluent with 13-23% of NH4-N, 15-21% of NO3-N, and 56-72% of NO2-N. Two cyclic modes with HRTs of 3 days (the ratio of aerobic feeding to total aerobic reaction time was 0.33) and 1.5 days (the ratio was 0.77) were employed. The cycle comparison between the two modes with different HRTs shows that there is no big difference with regard to the whole nitritation process, which is characterized by continuous conversion of loaded ammonium to nitrite and nitrate in the aerobic feeding period and no further conversion after the loading was terminated, resulting relatively stable levels for all the three nitrogen components in the entire cycle. Compared to 3-day HRT mode, 1.5-day HRT cyclic mode has doubled daily output in volume. In order to better understand the process of nitrite accumulation, more bench experiments were performed including an effluent nitrogen composition stability test and a reducing load test. The nitrite production stability was tested using four different ammonium loading rates, 0.075, 0.062, 0.053, and 0.039 gNH4/gMLSS*d in a 2-month running period. The TIN composition in the effluent was not affected when the ammonium load was between 0.053 and 0.075 g NH4/g MLSS*d (64% NO2-N, 16% NO3-N, and 20% NH4-N). Under 0.039 g NH4/g MLSS*d, a little more NO2-N was transformed to NO3-N with an effluent of 60% NO2-N, 20% NO3-N, and 20% NH4-N. The reducing load test has revealed the relationship between a declining FNA concentration and the decreasing nitrite production. The NH4+ load was gradually decreased from 0.081 to 0.011 g/gMLSS*d. When the NH4+ load was between 0.081 and 0.035 g/gMLSS*d, the ratio of NO2-/( NO2-+ NO3-) was kept stable around 0.75. When the NH4+ load dropped from 0.035 to 0.024 g/gMLSS*d, the ratio dropped to 0.70, accompanied by an abrupt decline of FNA from 1.2 to 0.6. From that point forward, the nitrite dominance environment in the system was no longer existing. Combining the results from both reducing load and stability tests, it is concluded that an ammonium loading rate around 0.035 is the lower threshold for producing a nitrite dominance effluent from the activated sludge SBR. In the denitrification step, three COD/NOx-N ratios (3.6, 4.8 and 6) and two solid retention times (SRTs), 16 and 23 days, were selected to test the influence of carbon availability and SRT on the total inorganic nitrogen (TIN) reduction and phosphorus removal efficiencies for the step-fed SBR. The best operating combination of parameters would consist of a COD/NOx-N ratio of 4.8 and an SRT of 23 days to achieve 97% TIN and 67% dissolved phosphorus (DP) removals.

Elucidating the oxidation mechanism of alloy has significant practical and theoretical impact. The addition of alloying elements could substantially alter the oxidation behavior of the base metal. The aims of this thesis research are to model alloy oxidation by investigating Cu-Au oxidation using an in situ ultra-high vacuum transmission electron microscopy (UHV-TEM). Cu-Au system was chosen because 1) Au does not form stable oxide at the reaction conditions, thus only Cu is expected to oxidize; and 2) Our extensive prior experiments of Cu nano-oxidation.The main findings of the research on Cu0.5Au0.5 (001) oxidation as compared to Cu (001) are: 1) segregation of gold atoms to the surface delays adsorption of oxygen and thus delays the nucleation (longer incubation time); 2) addition of gold enhances the nucleation by reducing nucleation activation energy (fast nucleation rate); 3) Oxide island growth is initially limited by surface oxygen diffusion. As oxide grows, a gold build-up zone forms around the oxide and hinders further growth by inhibits Cu supply to the reaction front; 4) Oxide grows slower due to the presence of gold.

This work mainly focused on fabricating of nanoscale polycrystalline organic and conjugated polymeric thin-film field-effect transistors and investigating their scaling behaviors of electrical transport and chemical sensing properties. Devices with channel lengths systematically ranging from a few hundred microns down to sub 10 nm were successfully fabricated with the techniques such as stencil mask, photolithography, electron beam lithography, and break junction. The use of a novel four-terminal geometry ensures that the active area for charge transport and vapor sensing is truly nanoscale, and eliminates undesirable spreading currents traveling over the large area outside the defined channel to reduce the background signal level. It was discovered that upon scaling channel lengths from micron scale down to nanoscale, the dominating factors for charge transport and vapor sensing in organic thin-film transistors become different. At small dimensions, injection limited transport and field-dependent mobility are the dominant mechanisms for transport through the gate-modulated channel at low and high longitudinal fields respectively. Furthermore for sub 10 nm channels, tunneling effect plays an important role. In micron scale devices, the drain current usually decreases as a sensing response upon exposure of the polycrystalline organic/polymeric semiconductor layer to the analyte, mainly because of the transistor threshold shift caused by the immobile charges at grain boundaries trapped by the dipolar analyte molecules. The vapor sensing behavior of nanoscale organic transistors is markedly different (in an opposite direction of response) from that of large-scale devices for the same analyte-semiconductor combination, due to the fact that the electrical transport in a nanoscale OTFT depends on its morphological structure and interface properties (such as the injection barrier at the metal-organic semiconductor contacts) which could be modulated by the delivery of analyte.