Search

We demonstrate a tunable optical filter using Ge 2 Sb 2 Se 4 Te 1 (GSST) phase change material to enable new capabilities in driver assistive technologies. The design principle of the optical filter is based on the metal-insulator-metal (MIM) coupled cavity theory. This filter selectively passes 850nm wavelength to enable daytime color and nighttime NIR imaging.

This paper describes a vision based deep learning approach to estimate the pose of a robot arm from a single camera input, without any depth information. Conventionally, pose of robot arm is determined using encoders which sense the j oint angles, and then the pose of each link (including the end effector) relative to the robot base is obtained from the direct kinematics of the manipulator. But there may be inaccuracies in the determined pose when the encoders or the manipulators are malfunctioning. This paper presents an approach based on computer vision, where a single RGB camera is fixed at a distance from the robot arm. Based on the kinematics of the manipulator and the calibrated camera, the input 2-dimensional image is reconstructed in 3-dimensional form and the pose of the manipulator is determined by means of a deep network model trained on synthetic data. Furthermore, a graphical user interface (GUI) is developed, which simplifies the output interpretation for users who operate the implemented system. Finally, the effectiveness of the proposed approach is demonstrated via several examples and results are presented. The proposed approach cannot entirely replace the function of encoders. Instead, it can be treated as a backup method which provides a reference solution.

In this paper we show that the internal quantum efficiency of NiSi Schottky-barrier photodetectors can be significantly improved as the silicide film thickness is reduced close to its percolation threshold. We fabricated photodetectors in two optical configurations (front-side and back-side illuminated) at four different film thicknesses between 1-4 nm as well as a thick (100 nm) reference device. We simultaneously fit the reflection and transmission data for each silicide film on silicon as well as for front and backside illuminations to extract the refractive index dispersion of each film. Using this technique, we can accurately determine the absorption of each constituent layer and extract the wavelength-dependent internal quantum efficiency from the external quantum efficiency. We show that the internal quantum efficiency is highly dependent on the silicide film thickness while the dark current is not. The internal quantum efficiency of our thinnest detector is the highest reported of any silicide Schottky-barrier photodetector of comparable barrier height to date with a 57× improvement over the thick (reference) device. Using an approximation to Vickers' model, we were able to fit the IQE spectra to extract the hot carrier mean-free path of electrons in NiSi.

As the technology costs of solar rooftops decline, small scale rooftops continue to become viable and grow without formal subsidies from the utility companies. Many new installations will continue to exist behind-the-meter (remain invisible), which poses challenges to the operating utilities for infrastructure and operations planning. This paper presents an approach to estimate the behind-the-meter solar PV size and the battery capacities. The proposed approach is based on energy balancing in buildings, and relies on data that is commonly available with utilities. The proposed approach is validated using a real-world dataset of 716 residential customers from a developed economy. The approach is compared with two baselines used by the studied Utility company. Results from the case study show that in 85% of the cases, the proposed approach has an accuracy of 98% in estimating the rooftop PV capacity (accuracies of baselines were in range 0%-30%). In estimating the battery capacity, the approach's estimates had less than 20% error in 70% of the cases (versus 35% error for baselines). The approach is also capable of discovering battery (dis)charging schedules, which is an additional useful information for utilities.

Electrical utilities mitigate temporary demand-supply mismatch by offering incentives to their customers to reduce electricity consumption. This is called demand response. While previous literature have addressed incentive design for demand response from the utility's perspective to improve their bottomline, they have ignored marketing expenses that consume a significant fraction of budget for a demand response program.In this paper, we optimize the marketing strategy for a residential demand response program by everaging the social network of the customers while also accounting for the electricity reduction potential of the customers. The problem is solved from the utility's perspective. Specifically, the utility has to decide the seed nodes and the value of referral reward for viral marketing given a budget. We propose an efficient solution technique for the problem and demonstrate effectiveness of the proposed technique.

The advancements in the power grid due to integration of new technology arises concerns regarding its reliability in terms of performance and security. On one hand, the gradual shift towards renewable energy sources leads to rise in uncertainty in terms of control and demand satisfaction. On the other hand, the integration of communication devices in order to make the grid smart increases its vulnerability to malicious activity. Automatic Generation Control (AGC) which is needed to maintain the system frequency and inter-area exchange has an important place in both the concerns. This paper presents an attack-defense analysis on the AGC operation of the power grid under varying conditions of renewables. It has two main contributions. First, a cyber-attack is conducted on the AGC algorithm with various levels of renewable penetration, to analyze the effect of an attack with renewables. Then a new algorithm for AGC using a PID control based approach is used and the attack is repeated to evaluate the impact. Secondly, an algorithm for attack mitigation is designed and its performance is analyzed with both the AGC algorithms. The various factors considered are effectiveness in reducing the impact on the system and adverse effects on normal operations. The experiments were conducted using the PowerCyber CPS security testbed at Iowa State University.

We investigate the thickness dependence of internal quantum efficiency for NiSi/n-Si Schottky barrier photodetectors. We observe a 20-fold improvement between the thinnest and thickest films tested and find that internal quantum efficiency improves until the film becomes discontinuous, falling below its percolation thickness.

The development of active metadevices continues to present keystone challenges in fields of plasmonics and photonics. Here, we demonstrate an analogue of electromagnetically induced transparency (EIT) effect in a far-infrared metasurface device via near-field coupling of bright and quasi-dark resonances resonating at nearly the same frequency with contrasting line widths. The proposed metasurface was further optimized numerically in order to demonstrate a reconfiguration effect (frequency-shift of the spectral response). The tunability property of the device is achieved by incorporating a thin layer of Ge 2 Sb 2 Te 5 (GST), a temperature-driven phase change material (PCM). Theoretical analysis based on a coupled Lorentz oscillator model explains the physical mechanism in the proposed design and shows a good agreement with the observed results. Such active hybrid EIT metadevices could have applications in tunable slow-light effects, delay bandwidth management and ultrafast laser induced switching.

There has been growing interest in developing multispectral imaging sensors to acquire more spectral information beyond the minimal metametric three-colors sensors [1]. This will enable chemical and material identification, and improve the accuracy of object recognition. Conventional multispectral imaging concepts are based on using an array of narrowband optical filters. In contrast, we have developed and demonstrated a novel Fourier multispectral imaging sensor which uses dielectric filters with different periods of sinusoidally varying transmittance. These are broad band filters, so unlike narrowband filters, very little signal is lost in the filters. The spectral resolution is determined by the number of sinusoidal filters — more filters result in a more precise spectral reconstruction. Each sinusoidal filter is based on a single cavity resonant structure, using ZnS as the cavity material on a fused silica substrate [2]. Initially, these Fourier filters were fabricated as bulk optical elements and integrated into a camera imaging system. Measurements were taken using a number of different LEDs as the targets. The data (Fig. 1) shows that spectral features such as transmission and absorption peaks are preserved very well, which makes this a versatile technique than narrowband filters for a wide range of multispectral imaging applications [3].

We report on the fabrication of silicon nanostructures using a metal-assisted chemical etching process. These vertically standing structures (we call quantum walls) could feature absorption in the infrared regime beyond the band edge of bulk silicon for potential application in infrared sensors.

Blocker tags are special-tags conceived to protect the customers privacy from unauthorized readers that viciously attempt to read the RFID tagged products carried by them. Unlike normal RFID tags, the blocker tag participates repeatedly in a tag-reading process and generates fake tag-IDentifiers (IDs). In all other respects it behaves well like a normal tag so that a reader invariably treats each fake ID as if it were a true ID from an actual tag. In other words, the blocker tag emulates multiple tags in the system. The reader's output is therefore obscured with an indistinguishable mix of fake-IDs and actual tag IDs. In an ALOHA-based tag reading system, a blocker tag probabilistically participates in the MAC protocol arbitrated by the reader. The higher the probability more contaminated the reader's result (with fake-IDs), and higher degree of protection from privacy snooping. Despite the salient use of the blocker tags, they can also be used maliciously to perform a Denial-of-Service attack on a trusted reader. As a result, the reader indefinitely interrogates the tags, and also incapable of isolating the actual tags in the system. While existing works detect and/or identify blocker tags in the system; to the best of our knowledge, there is no automated solution proposed to enable an authorized industry-compatible ALOHA-based RFID reader to perform a time-bounded effective tag-reading in the presence of a malicious blocker tag. In this paper, we present a simple yet novel solution to detect and jam the blocker tag from interrupting an on-going ALOHA-based tag-reading process. Through mathematical investigation and subsequent validation by simulation, we demonstrate the effectiveness of our solution.

Protocols are introduced to improve the tag reading efficiency of an RFID system consisting of a mobile reader that is constantly moving at a reasonable speed in a warehouse containing many thousands of items tagged with RFID. The warehouse is divided into storage areas with aisles separating these areas and the mobile reader tries to capture as many tag identifiers (IDs) as possible during this process. We execute ALOHA based protocols as the reader moves along the aisles and our goal is to read all the tags using a minimal number of laps, where a lap is the arrival of mobile reader at a location in the warehouse. To reduce the number of laps, we store the ALOHA protocols' status information (for example, the number of idle, collided, and successful slots) along with the location of the mobile reader at the mobile reader. During the next visit to the same location, the mobile reader uses the information (instead of executing the ALOHA based protocols from start) to improve the reading ability. We have further enhanced our protocol by segregating tags into zones (which is automatically done by the mobile reader) and allowing only tags within particular zones to participate in the ALOHA based protocols. This segregation results in lesser collisions and hence contributes to minimizing the number of laps required to complete the reading process. We have provided simulation results involving thousands of tags in a warehouse and the results show that our protocol reduces the number of laps by approximately 1/3rd in comparison with the standard ALOHA based protocols.

Municipal water delivery networks face increasing demands due to population growth. We focus on enhancing a water utility's infrastructure to meet its growing demands in a cost effective manner. Specifically, we consider the problem of placing pressure boosting pumps in a water network such that the minimum required delivery pressure is maintained at all the consumption points in the network for a given demand and the pump placement costs are minimized. The cost could be either energy cost or total cost dollars. Iterative optimization strategies and evolutionary computation techniques are typically used for solving such enhancement problems. We take a different perspective exploiting the structure of the network using graph theoretic principles. For water networks with tree topologies, we determine the optimal pump placement in terms of energy costs. We find that this energy optimal solution need not always minimize the total cost of ownership (TCO) involved in the pump placement. Therefore, we propose heuristic methodologies that reduce the TCO involved in placing pressure boosting pumps in tree networks. For water networks with complex topologies involving loops, we use the best of our tree solutions to find the initial seeds for iterative search strategies such as genetic algorithms (GA) and successive linear programming (SLP). We use EPANET for hydraulic modeling and study the efficacy of the proposed solution in terms of the TCO. In real-world topologies we considered, our heuristic seeding improves the performance of GA and SLP by about 68 % and 26 % respectively.

Data gathering is a typical operation in wireless sensor networks where data flow through a data gathering tree towards a sink node. Plenty of work had been done on constructing energy-efficient data gathering trees at the routing layer. However, data generation in wireless sensor networks could be bursty as it is driven by the events of interest. Without any sleep scheme, the sensor nodes will stay idle for most of the time. Therefore energy-efficient data gathering trees alone are not enough for energy saving. We propose a routing layer data gathering sleep scheme (DGSS) which could be incorporated into existing data gathering tree formation algorithms to save unnecessary energy consumption due to idle listening and meanwhile adapts to bursty traffic condition. We compare DGSS with DMAC, a popular energy efficient MAC protocol for data gathering, through simulations and show that DGSS can perform better than DMAC for bursty traffic at relatively high data rates.

We demonstrate an improved technique to grow glancing angle nanostructured thin films of silver and copper using electron-beam evaporation. Silver has particular interest for optical metamaterial activity, but unlike most other metals, the surface mobility of these atoms makes it difficult to grow distinct nanocolumns. By cryogenically cooling the substrate, we demonstrate that these metals can show marked improvement in their nanocolumn structure.

A framework for increasing throughput of RFID systems in an environment characterized with passive RFID tags on a mobile conveyor belt and multiple readers is introduced. The proposed framework involves information sharing among readers to improve the existing tag identification protocols in the mobile setting. The framework applies modifications to Tree and ALOHA protocols to reduce idle and collision times, thereby improving tag reading rates. The implementation of framework for two possible scenarios is described. Simulation results, whose performance aligns with the analytical models developed here, show that the framework improves the average tag reading rate of existing Tree and ALOHA based tag reading protocols.

As an instance of using IT to green non-IT domains, we consider the question whether lessons from computer networking can be applied in water distribution networks to improve their energy footprint and/or efficiency. Our contributions in this work are: (i) we identify several areas where principles from computer networking can be used to better water distribution; (ii) we focus on a specific infrastructure enhancement problem caused by increasing demands on a water utility network and present solutions (similar to those used in computer networks) that optimize both operational expenditure and total cost of ownership. We validate our solutions through simulations and compare their efficacy against techniques that are traditionally used in enhancing water networks. Our results show that lessons from computer networks can help in enhancing water networks.

There has been growing interest in adding real-time polarimetric imaging functionality to existing infrared focal plane arrays and CMOS/CCD imagers1. One approach is to fabricate the micropolarizers using wiregrids. Other approaches using bulk polarizers have also been tried2. The most challenging aspect is the fabrication of the wiregrid micropolarizer array, which requires patterning at the nanoscale. In this work, we examine a novel approach based on a well known process in CMOS technology. We have utilized the Damascene process for metallization to avoid some of the difficulties in patterning high-aspect ratio metal grids.

A simple near-unity power factor converter using two current control methods- the Average Current Control method (ACC) and Hysteresis Current Control method (HCC)-has been employed in this paper for a wind energy conversion system feeding simple loads. A simple stepped wind speed profile, comprising the rated wind speed and a higher wind speed, has been considered and the performance parameters of the unity power factor converter compared for these two current control methods. The wind turbine drives a Permanent Magnet Synchronous Generator (PMSG), which is interfaced with the near UPF converter feeding a resistive load. The performance parameters under study are the Total Harmonic Distortion (THD) of the input currents to the converter and the input Power Factor (PF) to the converter. Simulation has been performed in PSIM and Simulink employing SimCoupler for co-simulation and the results verify the efficacy of the converter in delivering satisfactory performance.

This paper presents the fabrication technique and results obtained for a titanium oxide based memristor design. A memristor wafer was developed that contains isolated devices as well as small memristor crossbar arrays (36 devices). The current-voltage relationship of the devices shows a memristive switching effect that is more apparent when the devices are negatively biased.

Consider a wireless sensor network where each node has $K$ radios$r_1,r_2,\cdots,r_K$ such that the one hop reachability distance(resp. energy expended) of (resp. by) radio $r_i$ is greater than that of $r_j$, $1 \leq j < i \leq K$. Given such a network, the problem of energy efficient radio activation is to minimize the total energy spent by the active radios across all nodes in order to maintain a connected network. We show that this problem is NP-Hard. We initially pay attention to the case of $K=2$ and discuss a basic version of the radio activation problem in such networks. We propose approximation methodologies for solving this problem. Our analytical and experimental studies reveal that the greedy algorithm and the minimum spanning tree solution have the best {\em worst case} performance while the greedy algorithm has the best {\em average case} performance. To the best of our knowledge, this is one of the first few works to focus on %deterministic solutions for optimal radio activation in generic multi-radio wireless networks.

The effects of Methyl-Tetra-Butyl-Ether (MTBE), as an additive, on the performance, combustion and the emission characteristics of a compression ignition (CI) engine are investigated experimentally. The experimental set-up consists of a diesel engine coupled with an eddy current dynamometer, AVL Di-gas analyser, AVL smoke meter and a data acquisition system is used. The di-gas analyzer is used to measure NO, CO and HC in the exhaust gases and the smoke meter is used to measure the smoke coefficient of the exhaust gases. The whole investigation is carried in two phases using the neat diesel in the first phase and the MTBE blended diesel fuels in the second phase. The MTBE blended diesel fuels are prepared in the varying proportions of MTBE (2, 4 and 6 %); subsequently the stability characteristics of the fuels are studied under static conditions. All the experiments are conducted at a constant speed of 1500 RPM by varying the brake load and the data obtained from the experiments are used to evaluate the performance and the combustion characteristics while, the emission characteristics are directly measured. The experimental results revealed that a significant improvement in the performance and appreciable reduction in the emissions for the MTBE blended diesel fuels on compared with neat diesel. The combustion analysis is supported the outcome on the performance and emission of the engine.

Wireless Sensor Networks (WSNs) open up wide range of possibilities for new applications such as monitoring and surveillance applications due to their low-cost and easy deployability. However, the limitations in terms of limited CPU processing power, memory, and battery, bring in a great deal of challenge to provide adequate security for on-board data storage and communication. While many works on secure key establishment exist, most of them assume that sensor nodes are either in fixed location or move within a single WSN. As WSN technology develops, there are more applications coming up which require a sensor node to roam across multiple WSNs. This paper presents a roaming protocol that provides security for the on-board sensory data (in terms of data access) while allowing a sensor node to move across multiple WSNs.

Cognitive radio has been proposed as an innovative and effective technology to exploit the efficient reuse of the precious radio spectrum. In spectrum sensing which is a crucial step in cognitive radio networks, an adequate sensing range needs to be defined to avoid the hidden transmitter/receiver problems. Moreover, a secondary user may have unreliable detection of the existent primary users due to multipath fading and/or shadowing. To solve these issues, we propose a cooperative spectrum sensing scheme with adaptive sensing ranges, especially for the hidden receiver problem. We explain the idea of using the adaptive sensing range at a secondary user transmitter and show with the simulation results that cooperative spectrum sensing with adaptive sensing ranges achieves a higher detection accuracy than that with a fixed sensing range.

Container packages experience diverse forms of mechanical excitations during their transportation and handling. Some of these excitations can damage the goods and make them reach their final destination with defects. Early detection of damages to container packages during their transport on a container truck can allow the truck driver to take the necessary steps to avert larger damages and thus cause significant cost savings. Sensor based approaches are being investigated to provide this early detection capability. This paper presents an approach that uses wireless vibration sensors based on Zigbee protocol to monitor the integrity and safety of packages during transportation. T-mote Sky wireless nodes integrated with two 2-axis MEMS accelerometers were used to monitor the integrity of the packages. Experiments were conducted to discern the vibration patterns resulting from some common modes of mechanical stability losses, such as wobbling, tilting, colliding and sliding. The experiments used a 1:32 scaled version of RC truck and a proportionately sized container. The vibration patterns under multiple stability loss modes were captured. It was observed that each type of stability loss can be clearly classified based on the patterns in the vibration data collected. Extraneous signal components were suppressed using a wavelet analysis, and fidelity of the signals capturing the pattern associated with the stability loss events was enhanced. Thus, each stability loss is associated with a specific set of abnormal (out-of-control) behavioral patterns exhibited by the processes vibration signals. An interrogation and detection procedure was developed based on this wavelet analysis to detect in realtime the stability loss as well as the times and identities of the specific modes of stability loss that occurred during the span of time over which the measured data is collected. The results show that the multi-scale monitoring facilitated by the wavelet analysis of signals from the wireless vibration sensor tags can be useful for accurate detection of stability loss events.

This paper outlines the design and development of a low-cost, small-footprint wireless sensor capable of sampling and transmitting sensory data at frequencies more than 6000 Hz. This sensor has been designed and developed in-house using off the shelf components and open source software for container integrity monitoring. The paper presents details of the various refinements that have been carried out to improve the default performance of the sensory unit. Laboratory tests indicate that while the developed sensor has comparatively higher power consumption than other commercially available sensory units, its small form factor and low cost make it attractive enough for several applications. The paper then describes how these sensors can be used to realize a reliable, round-the-clock system for monitoring the stability and integrity of cargo containers.

We can improve the end-to-end throughput between a sender and receiver in a wireless network using multiple paths which do not interfere with each other. Given finding such paths is computationally hard, we present a heuristic that is amenable to distributed implementation. This heuristic allows for the possibility of interfering links but schedules the transmission in ways to achieve the maximum possible throughput. One can characterize interfering links as either destructive or non-destructive, the latter of which does not affect the throughput, if transmission is properly scheduled. We present a distributed algorithm which exploits this knowledge to find multiple s-t paths such that high throughputs could be achieved despite the interference. On a wireless network with n nodes, our distributed algorithm has a time and message complexity of O(n2).

Data generation in wireless sensor networks could be bursty as it is dictated by the presence or absence of events of interest that generate these data. While conventional sensor nodes possessed only one radio interface, next generation sensor nodes are expected to have two (possibly more) radio interfaces, each with different range, capacity, and power consumption. Equipping sensor nodes with dual radios has its own benefits and can be quite useful in handling bursty traffic while at the same time satisfying the application's delivery requirements. In this paper, we propose an adaptive interface switch agent that intelligently selects the interface to be used for data transmission at a sensor node based on the data burst length while taking into consideration power consumption, throughput, and endto- end delay. The proposed work generalizes earlier works in this area to enable both the source nodes and intermediate data forwarding nodes to initiate the activation of high power radios so that they can be utilized to a higher degree for convergecast communication. We have performed extensive simulations with sensor nodes containing both IEEE 802.15.4 and IEEE 802.11 compatible radios. Our simulation results indicate that: (i) the end-to-end delay and throughput achieved by the proposed interface switch agent are comparable to those achieved in a network of sensor nodes equipped only with IEEE 802.11 radios, (ii) the energy consumed in the network using our interface switch agent is a fraction of that consumed in a network of the IEEE 802.11 sensor nodes and is comparable to that of sensors using only IEEE 802.15.4 radios.