Federated learning (FL) relies on a central authority to oversee and aggregate model updates contributed by multiple participating parties in the training process. This centralization of sensitive model updates naturally raises concerns about the trustworthiness of the central aggregation server, as well as the potential risks associated with server failures or breaches, which could result in loss and leaks of model updates. Moreover, recent attacks have demonstrated that, by obtaining the leaked model updates, malicious actors can even reconstruct substantial amounts of private data belonging to training participants. This underscores the critical necessity to rethink the existing FL system architecture to mitigate emerging attacks in the evolving threat landscape. One straightforward approach is to fortify the central aggregator with confidential computing (CC), which offers hardware-assisted protection for runtime computation and can be remotely verified for execution integrity. However, a growing number of security vulnerabilities have surfaced in tandem with the adoption of CC, indicating that depending solely on this singular defense may not provide the requisite resilience to thwart data leaks. To address the security challenges inherent in the centralized aggregation paradigm and enhance system resilience, we introduce DETA, an FL system architecture that employs a decentralized and trustworthy aggregation strategy with a defense-in-depth design. In DETA, FL parties locally divide and shuffle their model updates at the parameter level, creating random partitions designated for multiple aggregators, all of which are shielded within CC execution environments. Moreover, to accommodate the multi-aggregator FL ecosystem, we have implemented a two-phase authentication protocol that enables new parties to verify all CC-protected aggregators and establish secure channels to upstream their model updates. With DETA, model aggregation algorithms can function without any alterations. However, each aggregator is now oblivious to model architectures, possessing only a fragmented and shuffled view of each model update. This approach effectively mitigates attacks aimed at tampering with the aggregation process or exploiting leaked model updates, while also preserving training accuracy and minimizing performance overheads.

The relevance of angular stability (AS) and polarization independence (PI) is investigated by rotating the frequency selective reflectors (FSRs) kept behind a 10 GHz triangular slot antenna in orthogonal directions. Three distinct frequency selective surfaces (FSSs) based on closely packed rhombic loop (RL), square loop (SL) and merged dual rhombic loops (MDRL) are developed for the analysis under normal and oblique incidences. The gain enhancement of 4.87, 7.17 and 4.85 dB is achieved when the antenna is backed by 5 x 5 arrays of RL, MDRL and SL FSS, respectively. Later, with equal sized substrates, the gain enhancement of 7.25, 7.17 and 8.10 dB is achieved when the antenna is backed by 61 x 61 mm2 RL, MDRL and SL FSSs, respectively. It appears that stringent characteristics of FSSs i.e. AS and PI can be compromised for FSRs while employing antenna gain enhancement techniques.

A simple technique for gain enhancement of irregular hexagonal shaped ultra-wide band (UWB) slot antenna by inclusion of closely packed tessellated rhombic loop (TRL) based FSS is presented in this paper. The slot antenna is designed to cover entire UWB (3.1-10.6 GHz) frequency range with stable bi-directional radiation pattern. Also, it exhibits non-uniform boresight gain in zenith (theta = 0 degrees) and nadir (theta = 180 degrees) directions. The slot antenna is further loaded with TRL FSS at both slot and feed side separately to study distinct configurations of antenna integrated with FSS and evaluate the impact of the configurations of FSS on antenna performance. Later, an antenna and antenna loaded with TRL FSS prototype are developed for experimental validations. Through measurements it is found that the slot antenna covers entire UWB frequency band and provides a maximum of 4.31 dB gain in zenith direction while 4 dB gain in nadir direction. A gain enhancement of 2.44 dB in zenith and 3.8 dB in nadir directions are achieved when the slot antenna is integrated with TRL FSS along feed side and slot side separately. The measured results confirm the simulated results well.

The paper presents boresight gain enhancement of a microstrip fed X-band triangular slot antenna using rhombic loop (RL) based frequency selective surfaces (FSSs). Two single side printed orthogonal layers of arrays of dual rhombic loops (DRLs) kept in pairs are stacked together to form cascaded dual rhombic loop (CDRL) based reflective surface which exhibits wide stop bands within X-band but with reflection phase non-linearities. The non-linearities are suppressed when both layers are printed back to back (BTB) on both sides of same dielectric substrate while they are completely eliminated when both the layers are printed on same side of the dielectric substrate to form a merged dual rhombic loop (MDRL) based FSS. A triangular slot antenna operating at 8.2 GHz is designed and integrated with these FSSs for a purpose to investigate their impact on antenna performance. Through simulations, it is found that antenna when integrated with MDRL FSS demonstrates superior performance. Later, the slot antenna and MDRL FSS are developed and integrated to perform experiments. The integrated antenna exhibits an impedance bandwidth of 650 MHz and a maximum gain of 8.5 dB in antenna boresight including 4.5 dB enhancement due to frequency selective reflectors.

Conventional designs of arrays of metallo-dielectric frequency selective surface (FSS) for terahertz (THz) frequency range such as square conducting pads (Pixels), metallic mesh, mesh-loop (ML), square loop (SL) and complementary square loop (CSL) are analyzed and compared through simulations. Pixel FSS demonstrates 0.58 THz wide stop-band and 0.14 THz wide pass-band centered at 2.96 THz and 3.35 THz respectively. By complementing, the pixel array becomes a mesh array that appears as a band-pass FSS from 1.89 THz to 3.45 THz. ML array is formed by removing a square ring from the mesh array and provides a 0.32 THz wide pass-band and 0.3 THz stop-band centered at 0.82 THz and 1.15 THz. After eliminating the outer mesh of ML array, SL array yields a 0.32 THz wide band-stop FSS centered at 1.22 THz. Later, the CSL array is formed when SL array is complemented, and the array demonstrates a 0.6 THz wide pass-band at 1.14 THz. The pixel array provides a high signal rejection capability and the widest −10dB stop-band in the frequency range of 2.58-3.16 THz. This analysis and comparison is significant for reduction of EM coupling or interference in THz based integrated circuits.

Agriculture is an essential part of developing countries’ economies. Internet of Things (IoT) sensors can provide data about agricultural fields and then operate based on human input. The IoT is cloud-based technology that allows for scalable management, data management, data security, data analysis, and more. The present IoT solutions are not suitable for the Indian agriculture sector due to it being very complex and not being cost-effective, which means it is not applicable to the reality on the ground. The existing IoT solutions employ paid network GSM/NB-IoT/4G/5G for sensor networks. In this chapter, a holistic affordable IoT framework focused on the agriculture Indian scenario is proposed. The IoT end nodes are battery-operated, and power consumption should be low and adopt low-power wide-area network technology. Cloud services included the weather department, agriculture scientists of Indian Council of Agricultural Research, e-mandi, the Indian market, the nearest point for selling, and so on. It also includes the monitoring of temperature, humidity, soil moisture, irrigation and insect and pest detection, actuator intervention, message notification for disastrous weather warnings, and expert advice for farmers. Crop monitoring via by using machine learning and deep learning algorithms. In future work, this framework can adopt edge and fog computing to improve more secure data management.

This paper presents performance analysis of single layered bandstop frequency selective surfaces (FSSs) for C-band and ultra-wideband (UWB) frequency ranges respectively. These FSSs are based on polarization dependent (PD) single rectangular loop pair (RLP). RLP arrays in square and rectangular grids are developed and simulated to evaluate signal rejection and reflection for incident transverse electric (TE) and transverse magnetic (TM) waves. Further, orthogonal patterns are printed back-to-back (BTB) on dielectric substrate to study wave propagation and effect on wave polarization. Non-linearities in the reflection phase of BTB arrays are then mitigated by printing orthogonal FSS arrays on the same side of the dielectric substrate. The equivalent circuit models (ECMs) of these FSSs are also presented to predict their transmission characteristics. Later, two FSSs with wideband and ultra-wideband rejection capability are developed and experiments are carried out. The array of merged RLP arranged in square grid exhibits stopband from 4.5 GHz to 8.4 GHz with polarization independence (PI) and it is angularly stable while merged array of RLP arranged in rectangular grid exhibit ultra-wide stopband from 2.3 GHz to 11.5 GHz but it is not angularly stable. The measured transmission responses of all FSSs proposed here are close to simulated results.

We show that clock synchronized transmission has reduced receiver-side jitter that limits DMLs’ performance compared to conventional clock and data recovery. We access transmission quality using 35-GHz-bandwidth DMLs at 1271 and 1373 nm.

This paper presents performance analysis of Frequency selective reflecting surfaces (FSSs) based on rectangular, elliptical and rhombic loop unit elements respectively. Two orthogonally oriented configurations of polarization dependent (PD) arrays of rectangular loop, elliptical loop and rhombic loops are designed and their transmission characteristics under normal electromagnetic (EM) wave incidence are studied for comparison. These orthogonal PD arrays are first cascaded on both sides of the dielectric substrate as back-to-back (BTB) configurations and then merged and patterned on same side of the substrate. The transmission responses of both BTB and merged FSSs configurations are studied and compared. Later, all these FSSs are evaluated under oblique EM incidences. Although all FSSs maintain stable response even under oblique incidence in the microwave frequency bands, rectangular loop based FSSs provide excellent stability in their transmission responses. Merged FSSs outperform PD and BTB FSSs in the form of bandwidths (BWs) over which angular stable (AS) and polarization independent (PI) transmission responses are preserved. Merged rectangular loop (MRECT) FSS and Merged elliptical loop (MELIP) FSS remain AS and PI over 2.67 GHz and 1.65 GHz wide frequency bands respectively while Merged rhombic loop (MRL) FSS preserves AS and PI over 0.57 GHz wide frequency band.

An ultra wideband (UWB) slot antenna is proposed, developed, and experimentally validated in this article. The Norman window-shaped UWB slot antenna demonstrates stable radiation characteristics in both directions-zenith (theta = 0 degrees) and nadir (theta = 180 degrees) over the entire UWB frequency range (3.1-10.6 GHz) as observed in both simulations and experiments. A linear phase reflecting UWB frequency selective surface (FSS) is proposed by printing two orthogonal layers of pair of double rectangular loops on the same side of dielectric substrate as a merged frequency selective surface (MFSS). The UWB MFSS is then loaded with the slot antenna and the impact of MFSS on the impedance and radiation characteristics in different configurations of antenna and MFSS together is studied. One configuration with optimum solution is developed for experimental validations followed by the analysis of its radiation characteristics. The proposed antenna demonstrates a directional radiation pattern due to suppression of back lobe radiation when integrated with the UWB MFSS. The proposed antenna demonstrates maximum gain of 7.86 dB with a gain enhancement of 3.6 dB in boresight (theta = 0 degrees) direction when MFSS is applied as a reflector to opposite side of the antenna slot.

This work studied the co-pyrolysis of wheat straw (WS) and polyethylene (PE) via thermogravimetric experiments from room temperature to 1000 °C at various heating rates (10, 20, and 30 °C/min). Thermal behavior revealed that the maximum decomposition of WS, PE, and their blend occurred in three temperature ranges, viz. 250 - 496, 200 - 486, and 200 - 501 °C. Kinetic parameters were determined using model-free isoconversional methods. Activation energy from KAS (163.56, 220.26 and 196.78 kJ/mol for WS, PE, and blend), FWO (165.97, 222.05, 198.86 kJ/mol for WS, PE, and blend), and Starink (163.45, 220.05, 196.46 kJ/mol for WS, PE, and blend) method was estimated. From among various solid-state kinetic models, first-order reaction kinetics and one and two-dimensional diffusion models dominated co-pyrolysis of WS and PE. Thermodynamic parameters confirmed the feasibility of co-pyrolysis of WS and PE while differential thermal analysis signified that endothermic and exothermic reactions occur simultaneously.

Conventional square loop (SL) and complementary square loop (CSL) based frequency selective surfaces (FSSs) for distinct frequency range are designed, analyzed and compared in this paper. The SL and CSL type FSSs are designed to demonstrate band stop and band pass characteristics for S-band, Ku-Band, Ka-Band, G-band, Terahertz (THz) and Infrared (IR) frequency range respectively. The influence of dielectric substrate and metallic layer on FSS transmission characteristics are also studied for a purpose to identify the appropriate dielectric substrate and metallic layer to design the FSSs to obtain desired operating frequency, bandwidth of stop band and pass band and signal isolation/transmission. It is found that the signal transmission characteristics of FSSs at THz and IR frequencies deteriorates. Also, low lossy substrate materials are desired to improve the signal transmission and isolation levels at THz and IR frequency domains.

The progression of autonomic dysfunction from peripheral autonomic neuropathy (PAN) to cardiovascular autonomic neuropathy, including diabetic autonomic neuropathy and advanced autonomic dysfunction, increases morbidity and mortality risks. PAN is the earliest stage of autonomic neuropathy. It typically involves small fiber disorder and often is an early component. Small fiber disorder (SFD) is an inflammation of the C-nerve fibers. Currently, the most universally utilized diagnostic test for SFD as an indicator of PAN is galvanic skin response (GSR), as it is less invasive than skin biopsy. It is important to correlate a patient’s symptoms with several autonomic diagnostic tests so as not to treat patients with normal findings unnecessarily. At a large suburban northeastern United States (Sicklerville, NJ) autonomic clinic, 340 consecutive patients were tested with parasympathetic and sympathetic (P&S) monitoring (P&S Monitor 4.0; Physio PS, Inc., Atlanta, GA, USA) with cardiorespiratory analyses, and TMFlow (Omron Corp., Hoffman Estates, Chicago, IL, USA) with LD Technology sudomotor test (SweatC™). This is a prospective, nonrandomized, observational, population study. All patients were less than 60 y/o and were consecutively tested, analyzed and followed from February 2018 through May 2020. P&S Monitoring is based on cardiorespiratory analyses and SweatC™ sudomotor testing is based on GSR. Overall, regardless of the stage of autonomic neuropathy, SweatC™ and P&S Monitoring are in concordance for 306/340 (90.0%) of patients from this cohort. The result is an 89.4% negative predictive value of any P&S disorder if the sudomotor GSR test is negative and a positive predictive value of 90.4% if the sudomotor testing is positive. In detecting early stages of autonomic neuropathy, P&S Monitoring was equivalent to sudomotor testing with high sensitivity and specificity and high negative and positive predictive values. Therefore, either testing modality may be used to risk stratify patients with suspected autonomic dysfunction, including the earliest stages of PAN and SFD. Moreover, when these testing modalities were normal, their high negative predictive values aid in excluding an underlying autonomic nervous system dysfunction.

This paper reviews the complete framework of IoT-based agriculture systems with a specific emphasis agricultural sector of India. Agriculture is a major contributor to the economies of developing countries, and it assists with satisfying the fundamental necessities of food, pay, and work for the population. Minimal effort with low cost is a significant factor in making any IoT network valuable and satisfactory to farmers. LoRa is a fewer power consuming, long-range remote systems administration innovation, reasonable for small-rate large area of applications in the Internet of Things. An IoT-based agriculture framework provides better resource management, crop management, improved production quality and quantity, cost-effective farming, crop monitoring, and field tracking. The IoT equipment and networking methods related to wireless sensors work in agricultural applications are thoroughly investigated. Advance image processing methods and the Internet of Things together produce a new way of smart agriculture systems. IoT-based framework system includes monitoring of crop status, field irrigation, insect and pest detection on field, weather monitoring, actuator intervention, expert suggestions and warning system for farmers, and automations. This study aims to determine the better method for a low-cost IoT-based agriculture system.

This paper presents design and analysis of two band stop FSSs offering wider band rejection along with polarization independence and linear reflection phase characteristics separately. The polarization dependent vertical dual rhombic loop (VDRL) unit cells arranged with periodicity ratio 1:2 as well as horizontal dual rhombic loop (HDRL) with periodicity ratio 2:1 exhibit two separate stop bands for TE and TM incident EM waves. Both FSSs are evolved from orthogonally oriented polarization dependent arrays of rhombic loop pairs those are first printed on both sides of dielectric substrate back-to-back (BTB) to form a BTBDRL FSS and later merged together on same side of dielectric substrate to form merged dual rhombic loop (MDRL) FSS respectively. The transmission responses of both VDRL and HDRL FSSs for orthogonally polarized incident EM waves coincide when they are cascaded back-to-back on both sides of dielectric material. The non-linearities in their reflection phase responses are compensated by merging both the arrays on the same side of dielectric substrate. The array of BTBDRL exhibits 4.15 GHz wide stop band with two transmission nulls at 7.46 GHz and 9.52 GHz respectively while the MDRL FSS provides 4.8 GHz wide stop band centered at 8.2 GHz.

We demonstrate the first real-time 100 Gb/s IM/DD DMT transmission with a chirp managed laser over - 65.6 ps/nm to + 48.6 ps/nm of dispersion which is sufficient to support 400 Gb/s CWDM-4 transmission over 20 km.

We report on a comparison of different In-deposition schemes to achieve low areal densities of self-assembled InAs quantum dots (QDs) on GaAs(100) via the Stranski-Krastanov growth mode employing solid source molecular beam epitaxy. We could realize densities in the range of 10(7)-10(8) QDs/cm(2) utilizing homogeneous In deposition and an annealing step. At least on 70% of a 3 '' wafer the density was between 1 x 10(7 )and 1 x 10(8) QDs/cm(2). To achieve this, the In amount and the substrate temperature were controlled precisely. With inhomogeneous In-deposition via growing without sample rotation, we obtained low QD densities reproducible on a small fraction of the wafer surface. For a full-gradient, i.e., depositing the full In amount without rotation, the low-density area amounts in the best case to 10% of the overall wafer surface, whereas for a half-gradient, i.e. only half the In amount is deposited without rotation, it is 15%. The more In is deposited with substrate rotation, the less reproducible becomes the position of the low-density region on the wafer.

This paper reviews the high-speed directly modulated laser physics known for a long while yet recently have regained attention to realize a bandwidth reaching to 100 GHz. Chirp-managed laser for reach extension is also discussed.

Introduction. The severity and prevalence of Post-Acute COVID-19 Sequela (PACS) or long-COVID syndrome (long COVID) should not be a surprise. Long-COVID symptoms may be explained by oxidative stress and parasympathetic and sympathetic (P&S) dysfunction. This is a retrospective, hypothesis generating, outcomes study. Methods. From two suburban practices in northeastern United States, 152 long COVID patients were exposed to the following practices: (1) first, they were P&S tested (P&S Monitor 4.0; Physio PS, Inc., Atlanta, GA, USA) prior to being infected with COVID-19 due to other causes of autonomic dysfunction; (2) received a pre-COVID-19 follow-up P&S test after autonomic therapy; (3) then, they were infected with COVID-19; (4) P&S tested within three months of surviving the COVID-19 infection with long-COVID symptoms; and, finally, (5) post-COVID-19, follow-up P&S tested, again, after autonomic therapy. All the patients completed autonomic questionnaires with each test. This cohort included 88 females (57.8%), with an average age of 47.0 years (ranging from 14 to 79 years), and an average BMI of 26.9 #/in2. Results. More pre-COVID-19 patients presented with sympathetic withdrawal than parasympathetic excess. Post-COVID-19, these patients presented with this ratio reversed and, on average, 49.9% more autonomic symptoms than they did pre-COVID-19. Discussion. Both parasympathetic excess and sympathetic withdrawal are separate and treatable autonomic dysfunctions and autonomic treatment significantly reduces the prevalence of autonomic symptoms. Conclusion. SARS-CoV-2, via its oxidative stress, can lead to P&S dysfunction, which, in turn, affects the control and coordination of all systems throughout the whole body and may explain all of the symptoms of long-COVID syndrome. Autonomic therapy leads to positive outcomes and patient quality of life may be restored.