This paper presents a low-power and low-complexity data receiver circuit suitable for solar-cell-based optical receivers. The receiver circuit is based on a data slicer that has been modified to work while directly tied to a solar cell as the photo-transducer element. The receiver circuit uses a first-order RC high-pass filter to remove the solar cell’s DC voltage induced by ambient light and a comparator along with a voltage-averaging circuit to extract digital waveforms. A circuit analysis and experimental results are presented. The power consumption of the presented implementation of the receiver circuit is 6.28 μW and during tests, it achieved error-free transmission at a speed of 133 bps at a distance of 30 cm. This receiver circuit could be used as a wake-up receiver for solar-powered wireless devices or to transmit low-bandwidth command messages in visible-light optical communication applications.

A circuit for solar cell photo-luminescence modulation is presented. The circuit comprises a MOSFET, an opamp and a resistor forming a feedback loop that sets the current through the solar cell proportional to a modulating input voltage. The proposed modulator circuit achieves higher linearity than other photo-luminescence modulators making it suitable for direct amplitude modulation. A proof-of-concept prototype to demonstrate an optical link with a GaAs solar cell resulting in a bit error rate of 2.74×10-3using 6-ary pulse amplitude at a speed of 64.62 kbps.

This live demonstration showcases an optical communication system dubbed Optical Frequency Identification (OFID) that can be used to identify and track objects or animals as well as to sense environmental variables. The system consists of one or more optical tags and an optical reader. Each optical tag includes a GaAs solar cell, a low-power FPGA, an energy harvesting circuit and two sensors (light and temperature). GaAs solar cells are more efficient than their Si counterparts due to their direct bandgap electronic structure, which also makes them good emitters of infrared (IR) light when stimulated with light or with an electric current. This property, called luminescence, is exploited here to design electronic tags that use their on-board GaAs solar cells for three functions: 1) to transmit data (by modulating their IR luminescent emissions), 2) to harvest energy and 3) to receive information (by virtue of their photo-transduction property) [1] , [2] . The tags’ transmissions are received and decoded by the optical reader, which consists of a focusing lens, a photodiode, a trans-impedance amplifier followed by a gain stage and a digital decoder (implemented on an FPGA). Fig. 1 shows the live demonstration setup. It includes the optical tags, an optical reader, an oscilloscope to visualize received waveforms and a laptop to display received data packets.

This study explores the effectiveness of instrument identification algorithms (IIAs) using Convolutional Neural Networks (CNNs) and Random Forest (RF) models. Utilizing the NSynth dataset from Tensorflow, we began by evaluating the original CNN model, then enhanced its structure and feature selection to improve performance. The updated CNN model was compared to both the unmodified CNN and the RF model. The unmodified CNN achieved a practical validation accuracy of 55% and a training accuracy of 66%, indicating an 11% loss. The RF model demonstrated an average accuracy of 65%, with accuracies ranging from 92% for strings to 45% for reed instruments. Despite an initial inclination to refine the RF model, our research showed that CNNs could achieve higher accuracy with minimal preprocessing adjustments. By switching the input from mel spectrograms to MFCCs, we observed significant improvements. Our enhanced CNN model outperformed both the unmodified CNN and the RF model, highlighting the potential of CNNs in IIA development. However, this approach also resulted in increased training times and processing demands.

This paper presents a noise analysis of an optical receiver that uses a solar cell as a photo-detector to enable simultaneous energy harvesting and data reception. The noise analysis considers intrinsic noise generated by the electronic components of the receiver and includes the operation of a switching DC-DC converter, needed to draw the maximum amount of energy from the solar cell and transfer it to a load. The analysis yielded a noise model that is shown to be in good agreement with experimental measurements and was useful in identifying dominant noise sources, which in our case was a biasing circuit.

This paper presents a single-aperture, single-pixel reader for communication with Optical Frequency Identification (OFID) tags. OFID tags use solar cells to transmit and receive information wirelessly as well as to harvest radiant energy. Due to its single-aperture architecture, the reader's optical system provides a shared optical path for reception and transmission. Also, physical alignment between the reader and an OFID tag is visually guided using the reader's emitted light, securing a robust data link as long as the OFID tag is illuminated. In this paper, a description of the reader's optical and electronic sub-systems are presented. The transmitter and receiver circuits are described in detail. The transmitter, built with a linear LED driver, achieves a power efficiency of nearly 87%. The receiver, featuring a third-order bandpass filter, reduces both low-frequency and high-frequency ambient noise. A prototype of the reader was fabricated and housed in a custom 3D-printed enclosure. Test results show that the reader is able to receive modulated luminescent signals from an OFID tag at a distance of 1 m and at a data rate of 3 kbps.

This paper presents a photosynthetically active radiation (PAR) sensor based on an 18-band spectral sensor chipset with integrated optical filters. PAR sensors are an important tool for farmers and the proposed solution offers a path to low- cost PAR sensors. A technique for correcting the optical spectral leakage in the spectral sensor chipset is proposed and is applied to measurements collected indoors and outdoors. Tests results show that the proposed sensor and leakage-correction technique outperform lab-grade commercially-available PAR sensors.

Typical soil health assessment requires intensive field sampling and laboratory analysis. Although this approach yields accurate results, it can be costly and labor intensive and not suitable for continuous tracking of soil properties. Advances in soil sensor and wireless technologies are poised to replace physical sampling and offline measurement with in-field monitoring. This article reports the development, deployment, and validation of an Internet-of-Things (IoT) system for continuous monitoring of soil health. The end nodes of the proposed system, called soil health monitoring units (SHMUs), are solar powered and can be installed on a field for extended periods of time. Each SHMU transmits soil temperature, moisture, electrical conductivity, carbon dioxide (CO2), and geolocation data wirelessly using long-range wide-area network (LoRaWAN) radio technology. Data are received by a LoRaWAN gateway, which uploads it to a server for long-term storage and analysis. Users can view acquired data through a Web-based dashboard. The following significant experiments were carried out to validate the developed system: 1) a network consisting of eight SHMUs was deployed at an agricultural field site for several weeks and soil health metrics were analyzed using the soil health dashboard; 2) the flexibility of the system was demonstrated by the addition of an extra CO2sensor allowing an additional variable directly linked to soil health to be recorded; 3) a wireless communication range of 3422 m was estimated at a transmission power of 10 dBm by deploying the developed system on a large field; 4) the average current consumption of a SHMU (including its associated sensors) was estimated to be 13 mA, at this rate, the onboard Li-ion battery is able to sustain a SHMU for several days; and 5) a 7 cm ×6.5 cm solar panel was able to fully charge the onboard battery in 14 days while supplying power to the SHMU.

“Fail often to succeed sooner” is a common mantra that we are told is the secret to success. When reporting research results, however, scholars rarely write about their failed attempts and only focus on the successful ones. Perhaps the source of this disconnect between what we preach and what we do can be found in the underlying assumption that published work is meant to move the field forward and failed attempts supposedly do not. The goal of the confessions presented in this paper is to show that even failed attempts are genuine and valuable contributions to our field provided that we learn from our mistakes and correct them. The 27 confessions span from planning oversights, digital and analog design errors, misunderstanding of devices, overlooked parasitics, LVS errors, and troubles in testing.

An optical wireless communication approach that exploits the photo-luminescent radiation of LEDs is presented. In this approach the photo-luminescence of an array of LEDs is modulated by varying the impedance connected to the LEDs. The LEDs are also employed to harvest radiant energy making possible fully passive optical communications tags. Possible applications of this approach include short-range underwater communications. Initial experimental results suggest that communication speeds of few kilobits per second can be achieved.

This paper presents the design and development of a low-cost and low-profile sensor suitable for measuring Photosynthetic Active Radiation (PAR), which is defined as the photon flux density in 400 to 700 nm wavelength range. The sensor uses a silicon photodiode along with a bandpass optical filter and a diffuser housed inside a 3D-printed enclosure. A prototype of the sensor was built and tested with a halogen light source. Its performance was compared with research-grade commercially-available PAR sensors and with a spectroradiometer. It was found that the error of the sensor depends on light intensity being the most accurate when the light source was set to maximum intensity. The cost of the developed sensor is lower than commercial solutions, making it attractive for agricultural applications where cost is an important factor.

High-efficiency solar cells, such as GaAs solar cells, exhibit strong luminescent emissions in the infrared. This paper presents two circuits that are able to modulate these luminescent emissions while harvesting energy from the solar cell. These circuits can be used in Internet-of-Things applications where devices need an energy source and a means to transmit information wirelessly. The proposed circuits are based on a boost DC-DC converter and are suitable for binary (on-off) modulation. These circuits require only minimal additional hardware (either a switch or an AND gate) for their implementation. Proof-of-concept prototypes of these circuits were built and tested. Experimental results show a tradeoff between harvested energy and bit error rate.

A new wireless communication approach that employs solar cells as receivers and transmitters of information has been recently proposed [1,2]. This approach, which has been dubbed Optical Frequency Identification (OFID) for its analogy with Radio Frequency Identification (RFID), exploits the infrared luminescent radiation emitted by high-efficiency solar cells, such as GaAs solar cells, when excited by light (photo-luminescence or PL) or an electrical current (electroluminescence or EL). In OFID, the luminescent emissions of a solar cell are modulated by varying the voltage across the solar cell. Binary On-Off as well as multi-level modulations are possible in this scheme. The solar cell can also be used to receive information encoded optically by virtue of its photo-transduction property. Finally, in an OFID system the solar cell is also used for its intended purpose, which is to harvest radiant optical energy.

This paper presents a new solution for Internet-of-Things (IoT) applications in which GaAs solar cells are employed to transmit information. Transmission of information with a solar cell is possible by modulating the photo-luminescence and the electro-luminescence emissions of the solar cell. Photo-luminescence is modulated with switch to short circuit the solar cell terminals. Electro-luminescence modulation is achieved by applying a time-varying voltage to the cell. The solar cell can also be used to receive information encoded as light intensity variations. Moreover, the solar cell can also be used for its intended function of harvesting ambient radiant energy. Hence, a device equipped with a GaAs solar cell can reuse its on-board solar cell for multiple purposes resulting in devices with lower cost, size and complexity. The proposed devices, which are called Optical Frequency Identification (OFID) devices, are suitable for IoT applications in which wireless communications and energy harvesting are required features. A number of experiments are presented to demonstrate the validity of the proposed solution.

This live demonstration will show in practice how the photo-luminescence (PL) and the electro-luminescence (EL) emissions of a GaAs solar cell can be modulated to transmit information wirelessly.

In this paper, we present an energy efficient motion sensing platform for wireless body area networks that supports real-time, persistent gesture recognition through motion tracking sensors. The platform consists of multiple heterogeneous wearable sensors. The custom-built body sensors include inertial sensors, a low-power microcontroller and a 2.4 GHz radio transceiver. Signal classifiers such as Fishers linear discriminant classifiers, static neural networks, and focused time delay neural networks (fTDNN) are employed to classify the signals obtained from the wearable sensors. It was found that at a sampling rate of 10 Hz and just 4 bits/sample, the fTDNN classifier achieves 88% classification rate. Our results show that reducing the sampling and quantization rates could be used in energy constrained sensor networks for gesture recognition.

This paper presents a configurable image sensor with energy harvesting capabilities. The image sensor's pixels can be configured as photo-sensors for image acquisition or as tiny solar cells for energy harvesting. The proposed configurable pixel design requires four pixels and uses a N-well/P-sub photodiode which is typically regarded as a parasitic diode for energy harvesting. A chip containing a 64×64 array of configurable pixels, read-out circuitry and biasing was designed and fabricated on a standard 0.5 μm CMOS process. The pixel size is 36 μm×40.8 μm with a fill factor of 42%. Test results are presented demonstrating the dual functionality of the image sensor. Measurements show that 55 nW can be harvested for an illuminance of 500 lux at the lens plane.

Under the compressive sensing theoretical framework a sparse signal can be acquired using few random measurements. This result implies that an analog signal can be compressed while it is being acquired. However, compressive sensing does not yet achieve the high compression rates obtained with standard data compression techniques. To improve the compression performance of compressive sensing, the measurements can be further encoded exploiting their statistical structure. This work explores the concept of encoding compressive sensing measurements using a low-complexity entropy encoder such as the Golomb-Rice encoder. It is found, through system-level numerical simulations, that a Golomb-Rice encoder can reduce the bitrate of compressive sensing by more than 1 bps. Balanced and unbalanced sensing matrices were used in the simulations. Balanced sensing matrices resulted in a slightly better average SNR and CR performance.

This work evaluates the concept of employing light-emitting diodes (LEDs) to harvest solar energy. Due to the optical properties of the semiconductor materials employed in the fabrication of LEDs, LEDs can absorb photons and generate electron-holes. Thus, in principle LEDs can behave as small solar cells from which power can be harvested. Unlike solar cells, LEDs are not optimized to absorb light. As a result LEDs have lower light-to-electricity conversion efficiencies. However, there are scenarios such as outdoor displays or billboards where a large amount of LEDs are already installed. These LEDs represent an energy resource whose installation and fabrication costs have already been covered by the primary application of message displaying. With this motivation, an evaluation of the energy-harvesting capability of a number of LEDs has been carried out in this work. In the proposed concept, LEDs become a source and a sink of energy. A bidirectional buck-boost DC-DC converter able to transfer energy to and from the LED has been built and tested. It is found that red LEDs provide the highest power. Up to 2.76 W can be harvested from a 96×216 display.

The problem of signal recovery from non-ideally quantized linear measurements is considered. Quantization non-idealities due to circuit second-order effects are inevitable in practical deployments of compressive sensing and introduce distortion in the measurement process. Quantization non-idealities are commonly characterized by integral non-linearity (INL), offset and gain error metrics. These metrics are included in the signal reconstruction process to enforce quantization consistency. Signal reconstruction is formulated as a linear program. The performance of the proposed approach is assessed numerically and compared with other signal recovery techniques. It is shown that a linear program can be competitive and in some cases superior to more elaborate signal recovery approaches. It is also shown that satisfying quantization consistency does not always lead to better signal recovery in terms of signal-to-noise ratio (SNR).

The design and test of an imager that can perform the tasks of sensing and energy harvesting is presented. The imager is based on hybrid pixels whose photodiodes that can be configured to work as photosensors or tiny solar cells. The hybrid pixels allow the imager to harvest energy in between frames or while it is in sleep mode. As a proof of concept a 32 × 16 array of hybrid pixels has been designed and fabricated in a standard 0.5 μm CMOS process. A charge pump to boost the inherently low voltage of the photodiodes when they work as solar cells was also designed. The charge pump employs flying capacitors to maximize the amount of energy that can be harvested. Measurement results show that up to 3 μW of power can be harvested from the array under indoor illumination conditions. Images acquired with the imager are presented.

An image compression algorithm suitable for focal plane integration and its hardware implementation are presented. In this approach an image is progressively decomposed into images of lower resolution. The low resolution images are then used as the predictors of the higher resolution images. The prediction residuals are entropy encoded and compressed. This compression approach can provide lossless or lossy compression and the resulting bitstream is a fully embedded code. A switched-capacitor circuit is proposed to implement the required operations. A prototype has been implemented on a 0.5 μm CMOS process. Simulation and measurements results validating the proposed approach are reported.

We used Fisher linear discriminant analysis (LDA), static neural networks (NN), and focused time delay neural networks (TDNN) for gesture recognition. Gestures were collected in form of acceleration signals along three axes from six participants. A sports watch containing a 3-axis accelerometer, was worn by the users, who performed four gestures. Each gesture was performed for ten seconds, at the speed of one gesture per second. User-dependent and user-independent k-fold cross validations were carried out to measure the classifier performance. Using first and second order statistical descriptors of acceleration signals from validation datasets, LDA and NN classifiers were able to recognize the gestures at an average rate of 86% and 97% (user-dependent) and 89% and 85% (user-independent), respectively. TDNNs proved to be the best, achieving near perfect classification rates both for user-dependent and user-independent scenarios, while operating directly on the acceleration signals alleviating the need for explicit feature extraction.

An analog-to-digital converter suitable for compressive sensing applications is presented. The converter is based on an incremental sigma-delta converter architecture and is able to directly acquire and convert to digital format compressive sensing measurements. A switched-capacitor circuit is proposed for the converter's hardware implementation. Simulations at the system and transistor levels validate the approach. The converter occupies a small area of 0.047 mm2on a target 0.5 μm CMOS process. Thus, several of them can be implemented in parallel to achieve high conversion rates.

This paper presents the design and measurements of a predictive coding on-sensor compression CMOS imager. Predictive coding is employed to decorrelate the image. The prediction operations are performed in the analog domain to avoid quantization noise and to decrease the area complexity of the circuit. The decorrelated image is encoded with a bank of column-parallel entropy encoders. Each encoder is combined with a single-slope analog-to-digital converter (ADC) to reduce area complexity and power consumption. The area savings resulting from such combination allow to integrate an ADC and an entropy encoder at the column level. A prototype chip was fabricated in a 0.35 μm CMOS process. The output of the chip is a compressed bit stream. The test chip occupies a silicon area of 2.60 mm × 5.96 mm which includes an 80 × 44 APS array. Tests of the fabricated chip demonstrate the validity of the design.

In this paper, a focal plane SPIHT image compression scheme has been integrated on the focal plane of CMOS image sensor. A test chip is prototyped in 0.35μm technology to verify the design. In our image compression scheme, focal plane prediction replaces the traditional wavelet transform for image decomposition, and the traditional SPIHT coding is modified to reduce computational complexity and to suit the parallel implementation on the sensor focal plane. Simulated compression performance is still good, especially at low bit rates, compared to traditional SPIHT and other compression schemes. Integrating SPIHT compression on the focal plane not only reduces the sensor readout volume, but also provides all the advantages of SPIHT compression including both high image quality and optimized data ordering for progressive image transmission.

A CMOS image sensor has been designed to perform the front-end image decomposition in a prediction-SPIHT image compression scheme. The prediction circuitry based on charge sharing is integrated inside the sensor array to perform 3-level image decomposition. A CMOS test chip has been prototyped and tested. The test results justify the pixel design and demonstrate that lossy prediction based focal plane image compression can be realized inside the sensor pixel array to achieve a high frame rate but with much less data readout volume. Also, the sensor can be used to achieve comparable compression performance with much lower computational complexity compared to 2D discrete wavelet transform (DWT) based image compression.

A focal plane video compression integrated circuit is presented. The design consists of a 128 times 128 pixel array and a bank of column-level processors. Each one of the column-level processors performs the tasks of image decorrelation, quantization, and entropy encoding. The chip provides at its output a compressed bit stream. The integration of the quantizer and the entropy encoder at the column level is possible by sharing circuitry between a single-slope analog-to-digital converter and a Golomb-Rice entropy encoder. In addition, the design includes a low-complexity algorithm for the adaptation of the Golomb-Rice coder to the statistics of the video signal. The design has been fully verified through simulations and has been implemented in a 0.35 mum CMOS technology. The chip layout occupies an area of 7 times 5 mm2

We present a CMOS image sensor that performs focal plane image decomposition based on charge sharing computation circuitry. The effect of parasitic capacitance between capacitor bottom plate and substrate on the computational accuracy is discussed and a new circuit is also proposed to characterize the parasitic effects. The test results demonstrate that prediction based focal plane image compression can be realized inside the sensor array resulting in high compression performance using frame rate pixel parallel computation. This architecture can subsequently be combined with backend level testing encoding to form a complete compression sensor.