Underwater object detection presents significant challenges due to the intrinsic properties of light in aquatic environments. State-of-the-art methods often fail to capture the subtle details necessary for accurate detection in these scenarios. Recent advancements have shown promising results by reformulating relationships in graph space; however, most SOTA methods typically employ graph structures that are insufficient to represent the complex latent variables inherent in underwater environments. Hence, these models are unable to preserve the actual boundaries of object detection. To address these limitations, in this paper, we propose a novel end-to-end architecture that uses a graph refactoring aware deep learning based encoder-decoder architecture. The proposed approach uses a convolutional backbone to project the image into latent space, where an unsupervised initial graph is constructed. The hypergraph convolution is then utilized to optimize message passing between graph nodes, enhancing the representation of complex relationships of latent space. This helps in the retention of intricate details by modelling two or more latent variables as hyperedge by sharing the information among themselves. Finally, an image generation module maps the enhanced graph representation back to image space. The effectiveness of the proposed method is demonstrated through a comparative analysis against fourteen state-of-the-art methods on the benchmark underwater databases. The code for the proposed scheme can be found at https://github.com/immkapoor/hyper_graph.

Recently, Test-Time Adaptation (TTA) algorithms have gained traction in image/video quality assessment (IQA/VQA). These methods use virtual losses, like group contrastive and rank loss, as auxiliary tasks to adapt batch normalization parameters, helping models generalize better to distribution shifts between training and testing datasets. Group contrastive loss clusters images into low- and high-quality groups based on predicted quality scores, maximizing feature space distance between them. However, its effectiveness relies on the base model’s ability to accurately predict quality scores, which can be compromised when distribution shifts occur, leading to suboptimal adaptation and degraded performance. We propose a novel clustering approach based on the assumption that high-quality images contain richer high-level information, which is extracted using a pre-trained VGG-16 model. Images are clustered by comparing the VGG-16 features of the highest quality image in a batch with the others, enabling effective grouping based on feature affinities. These accurate clusters enhance the computation of contrastive loss, improving the adaptation of batch normalization layers. Additionally, we introduce an adaptive rank loss to reduce the impact of rank loss when the base model can distinguish images with varying distortion levels. Experimental results across multiple image quality assessment datasets, including LIVE, CID-2013, KONIQ-10K, and SPAQ, as well as algorithms like MetaIQA, HyperIQA, TReS, and MUSIQ, show that the proposed method consistently performs better than the existing Test-Time Adaptation (TTA) approach.

In this paper, a joint optimal allocation of transmit power at the source and jamming power at the destination is proposed to maximize the average secrecy energy efficiency (SEE) of a wireless network within a finite time duration. The destination transmits the jamming signal to improve secrecy by utilizing full-duplex capability. The source and destination both have energy harvesting (EH) capability with limited battery capacity. Due to the Markov nature of the system, the problem is formulated as a finite-horizon reinforcement learning (RL) problem. We propose the finite-horizon joint power allocation (FHJPA) algorithm for the finite-horizon RL problem and compare it with a low-complexity greedy algorithm (GA). An infinite-horizon joint power allocation (IHJPA) algorithm is also proposed for the corresponding infinite-horizon problem. A comparative analysis of these algorithms is carried out in terms of SEE, expected total transmitted secure bits, and computational complexity. The results show that the FHJPA algorithm outperforms the GA and IHJPA algorithms due to its appropriate modelling in finite horizon transmission. When the source node battery has sufficient energy, the GA can yield performance close to the FHJPA algorithm despite its low-complexity. When the transmission time horizon increases, the accuracy of the infinite-horizon model improves, resulting in a reduced performance gap between FHJPA and IHJPA algorithms. The computational time comparison shows that the FHJPA algorithm takes 16.6 percent less time than the IHJPA algorithm.

International audience

This work investigates the outage behavior of an uplink non-orthogonal multiple access (NOMA) based mixed underwater optical wireless-RF communication (UOW-RFC) system under dynamic power allocation (DPA) scheme. In particular, we utilize an improved dynamic power allocation (IDPA) scheme over underwater subsystem based on the quality-of-service (QoS) of users and employ partial relaying scheme (PRS) over the RF hop. Moreover, we consider fixed as well as optimum power allocation schemes over the relaying phase. Considering the mixture-Extended Generalized Gamma (mEGG) distribution with pointing error for underwater links and Nakagami-m distribution for RF link, we obtain the expressions for end-to-end outage probability for IDPA scheme in terms of bivariate Fox-H function. It is shown through the numerical results that, in comparison with the conventional DPA (CDPA) and Fixed power allocation (FPA) schemes, IDPA provides significant improvement in the outage performance of both high-QoS and low-QoS users. The numerical results also reveal that employing IDPA enables to overcome the error floor problem of uplink NOMA. The analytical results are corroborated through simulations.

This paper considers an optical reconfigurable intelligent surface (ORIS)-assisted free space optical (FSO) communication network to provide internet connectivity to the high speed train (HST). To maintain reliable FSO communication and compensate for severe pointing losses caused due to misalignment, we utilize a multi-aperture HST receiver that performs selection combining while achieving spatial diversity. In particular, we analyze the second-order statistics (SOS) at the HST receiver in the presence of atmospheric turbulence, foggy conditions, and misalignment error. We specifically develop the closed-form analytical expressions for the level crossing rate (LCR) and the average fade duration (AFD) of the system using the signal-to-noise ratio stochastic process under non-isotropic scattering environment. The effect of various system and channel parameters, including number of ORIS elements, number of receiving apertures, speed of the HST, misalignment error and fog parameters, mean angle-of-arrival, and degree of non-isotropic scattering, has been demonstrated on the LCR and AFD performances of the considered network. Further, by using finite-state Markov channel (FSMC) model, we derive packet error rate (PER) using the derived LCR expression. In order to analytically evaluate the optimum packet length for the FSMC model, we employ stop-and-wait automatic repeat request (SW-ARQ) protocol and maximize the system throughput. Additionally, it is shown numerically how different system settings affect the PER and throughput performances of the FSO-based HST communication network.

In this work, we consider the use of a simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) to assist the communication between the access point (AP) and two users, along with the existence of direct links. Considering the non-orthogonal multiple access (NOMA) based transmission by AP, we derive the expressions for coverage probability of both the users assuming Nakagami-m fading channels with the possibility of imperfect successive interference cancellation (SIC) at the near user. We also find the optimal values of NOMA power allocation parameter and STAR-RIS transmission coefficient that jointly minimizes the absolute difference in the throughput of the two users. The superiority of the proposed system is shown through numerical results in contrast to a STAR-RIS assisted orthogonal multiple access (OMA) transmission and a conventional NOMA scheme. The impact of various other system parameters like number of reflecting elements and fraction of imperfect SIC is demonstrated on the system performance.

In this paper, we explore the use of reconfigurable intelligent surface (RIS) in unmanned aerial vehicle (UAV) based multiuser downlink communications, where a flying UAV serves multiple single antenna users through multiple RISs mounted on various buildings. More specifically, we consider the selection of RISs based on the outdated and imperfect channel state information (CSI) of the composite UAV-RIS-User channels at the UAV. After selection process, the UAV communicates to the user via the selected RISs and also with the direct link. Particularly, we derive an infinite series based expression for selection probability of RISs under both the outdated and imperfect CSI of composite channels based selection scheme. We also derive the statistical distribution of instantaneously received signal-to-noise ratio (SNR) under outdated and imperfect CSI conditions of both the direct and composite links at the user. Next, using the derived statistics, we analyze the network’s performance in terms of the average coverage probability (ACP) and average bit error rate (ABER) over the complete UAV flight time. Moreover, we discuss the behavior of ACP and ABER for very small and very large values of UAV transmit power, respectively. It is depicted through numerical results that selecting more RISs from a group of small-sized RISs may not be as advantageous as selecting fewer RISs from a group of large-sized RISs. Moreover, we also demonstrate the effect of several system parameters such as number of RIS reflecting elements, number of selected RISs, the severity of UAV-RIS and RIS-User links, and the severity of imperfect and outdated CSI on the network’s performance. The analytical results are corroborated with Monte-Carlo simulations.

A new avatar of Ultra-Wideband (UWB) has emerged in recent times that provides users with reliable ranging measurements. In this paper we present an evaluation framework to ascertain the effectiveness of UWB for in-home device control, and help in improving the user experience. In the context of in-home device control, we chose a set of three use cases to explore challenges and evaluate value of using UWB for detecting the user intent based on the distance and angle measurements obtained by performing ranging. In this paper, we describe in detail the system architecture of our evaluation framework and the main component of the framework - User intent detection algorithm designed to improve the efficacy of UWB ranging technology for in-home device control. The user intent detection algorithm allows plugging in various kinds of filters and feedback/tuning mechanisms. We conducted evaluation of our user intent detection algorithm by plugging in Kalman filter and moving average filter, typically used in indoor positioning systems. We demonstrated the chosen use cases in lab setup and affirmed the effectiveness of UWB for in-home device control. Results reveal that our user intent detection algorithm reduces false positives considerably compared to plain UWB ranging measurements for triggering action and improves the effectiveness of UWB making the user experience more intuitive.

Reconfigurable intelligent surface (RIS) has emerged as an important transmission technology for numerous applications in Internet of Things (IoT) systems. Thus, in this paper, we investigate the application of RIS in energy harvesting (EH) based unmanned aerial vehicle (UAV) communication network with finite block length (BL) codes, where a rotary wing type flying UAV communicates with the multiple single antenna IoT users with the aid of multiple RISs mounted on several skyscraper buildings. To transmit the signal to a particular IoT user, the UAV selects an RIS on the basis of either UAV-RIS (i.e., partial) or UAV-RIS-IoT (i.e., full) channel state information (CSI) and then transmits the signal through the selected RIS along with the direct link transmission. In particular, we derive (i) the expression for probability of RIS selection, (ii) the statistical distribution of instantaneously received information signal-to-noise ratio (SNR) at the IoT user. Based on the derived statistics, we analyze the performance of the considered system under finite BL codes in terms of the average outage probability, average block error rate (ABLER) and goodput averaged over entire flying duration. Moreover, the BLER performance with finite BL codes is also compared with the infinite BL codes scenario. Additionally, we also investigate the impact of various channel and system parameters like imperfect CSI, number of RISs and the number of reflecting elements at each RIS, location of IoT users, variable altitude of the UAV, and the severity of channel fading of UAV-RIS link on the system performance. Furthermore, we have obtained the optimum UAV location in each time slot which minimizes the ABLER per time slot over all the users in the network. The analytical results are corroborated with Monte Carlo simulations.

Owing to the factors such as controllable mobility, ready-to-use technology, low cost, easy implementation, and so on, unmanned aerial vehicle (UAV) possesses tremendous potential to be one of the primary candidates for next-generation (6G) wireless networks. This paper presents a UAV-assisted multiuser communication network where a multiple antenna UAV base station (BS) serves multiple single antenna ground users (GUs). UAV-BS uses a laser source-based charging mechanism to fulfill its power requirement and applies simultaneous wireless information and power transfer (SWIPT) in the downlink in order to provide desired power to energy-constrained GUs. Also, a clustering-based hybrid multiple access technique is used that combines both orthogonal multiple access (OMA) and non-orthogonal multiple access (NOMA) signaling to transmit the information to all GUs, simultaneously. Due to the involved analytical complexity corresponding to the multiple antennas and users, we use a hybrid beamforming method for efficient communication. Next, we analyze the performance of the proposed framework in terms of user outage probabilities, their respective throughput, and average power harvested considering non-linear energy harvesting and derive expressions of these performance metrics. Moreover, we formulate an optimization problem where the throughput of one GU is maximized by optimally choosing the power allocation parameter while ensuring the desired target throughput at other GU in each cluster. We also illustrate how crucial is the optimal selection of the target rates to maximize the network performance. Simulation results are provided to validate the accuracy of derived expressions and to highlight the dominance of hybrid beamforming and hybrid NOMA compared to conventional methods on the performance of the considered network.

Reconfigurable intelligent surface (RIS) has emerged as important transmission technology to improve the spectral/energy efficiency in the next-generation (beyond 5G (B5G) and 6G) wireless communication network and has numerous applications in the areas of Internet of Things (IoT) and vehicular communication systems. Thus, in this paper, we investigate the application of RIS in unmanned aerial vehicle (UAV) enabled vehicular communication system with infinite and finite block length codes, where UAV communicates with the single antenna ground vehicle in the presence of several interfering vehicles on the road. We have obtained the approximate closed-form statistics of received SINR at ground vehicle in the presence of multiple nonidentical interference links. Furthermore, we analyze the performance of the considered system in terms of the coverage probability, bit-error-rate, block error rate (BLER) and goodput. It has been shown through the numerical results that the deployment of RIS significantly improves the performance of UAV-enabled vehicular communication network, even in the presence of the direct link between the UAV and the ground vehicle. Additionally, we also investigate the impact of various channel and system parameters like practical reflection coefficients of RIS, number of RIS reflecting elements, and number of interfering vehicles on the system performance. The analytical results are corroborated with Monte Carlo simulations.

Intelligent reflecting surface (IRS) is being considered as a prospective candidate for next generation wireless communication due to its ability to significantly improve coverage and spectral efficiency by controlling the propagation environment. One of the ways IRS increases spectral efficiency is by adjusting phase shifts to perform passive beamforming. In this letter, we integrate the concept of IRS aided communication to the domain of multi-direction beamforming, whereby multiple receive antennas are selected to convey more information bits than existing spatial modulation (SM) techniques at any specific time. To complement this system, we also propose a successive signal detection (SSD) technique at the receiver. Numerical results show that the proposed design is able to improve the average successful bits transmitted (ASBT) by the system, which outperforms other state-of-the-art methods proposed in literature.

In this paper, we apply the long short-term memory (LSTM), an advanced recurrent neural network based machine learning (ML) technique, to the problem of transmitter selection (TS) for secrecy in an underlay small-cell cognitive radio network with unreliable backhaul connections. The cognitive communication scenario under consideration has a secondary small-cell network that shares the same spectrum of the primary network with an agreement to always maintain a desired outage probability constraint in the primary network. Due to the interference from the secondary transmitter common to all primary transmissions, the secrecy rates for the different transmitters are correlated. LSTM exploits this correlation and matches the performance of the conventional technique when the number of transmitters is small. As the number grows, the performance degrades in the same manner as other ML techniques such as support vector machine, k-nearest neighbors, naive Bayes, and deep neural network. However, LSTM still significantly outperforms these techniques in misclassification ratio and secrecy outage probability. It also reduces the feedback overhead against conventional TS.

This study investigates the unified performance of a free space optical link with a finite‐sized receiver over an exponentiated Weibull distributed atmospheric turbulence in the presence of a misalignment error (ME). The unification is done for two detection techniques namely heterodyne and intensity modulation/direct detection. First, the unified expressions for probability density function, cumulative distribution function, moment generating function, and moments of instantaneously received signal‐to‐noise ratios (SNRs) at the receiver, have been derived. By utilising the derived statistics, the unified novel expressions for outage probability, average error probability (AEP), and ergodic capacity are obtained. Furthermore, the AEP performance is examined at a high SNR in order to obtain the diversity order and coding gain, analytically. The derived results clearly demonstrate the impact of severe atmospheric conditions, ME, aperture size, and type of detection scheme on the system performance. The derived analytical results are verified through simulations.

In this study, the performance of decode‐and‐forward relay‐assisted free‐space‐optical (FSO) communication systems under atmospheric turbulence‐induced fading and misalignment errors is investigated. To mitigate the adverse effects of the atmospheric turbulence, the aperture‐averaging receivers are considered both at the relay and destination sides. The atmospheric turbulence‐induced fading is modelled via the exponentiated‐Weibull distribution, which has recently been proposed to characterise an FSO link in the presence of finite‐sized receiver aperture. The expression for the moment generating function (MGF) of the instantaneous signal‐to‐noise ratio is derived. Furthermore, new closed‐form expression for the outage probability is obtained. Moreover, the new expression for the average symbol error rate of the subcarrier intensity‐modulated M ‐ary phase‐shift keying is obtained using the MGF‐based approach. Finally, numerical examples are discussed and all the derived analytical results are corroborated by Monte Carlo simulations.

The authors analyse a dual‐hop mixed radio frequency/free space optical (RF/FSO) communication system comprising of single‐input multiple‐output (SIMO) RF hop and a single FSO hop. The RF source is connected to an optical destination through a decode‐and‐forward relay having RF/FSO capabilities. Each link in SIMO‐RF hop is assumed to experience independent and identically distributed (i.i.d.) Nakagami‐ m fading. For FSO hop, a unified novel expression for the probability density function (PDF) of irradiance has been derived which unifies Gamma–Gamma and generalised‐ K ( ) distributions with misalignment error. Moreover, we obtain a unified PDF of instantaneous signal‐to‐noise ratio (SNR) by unifying the heterodyne detection and intensity modulation/direct detection schemes. Utilising the unified PDF, we derive novel unified closed‐form expressions for average symbol error probability (for different RF modulation techniques), outage probability, and ergodic capacity for the system under consideration. We have also analysed the considered system for high SNR conditions and have analytically obtained the unified diversity order of the same. The derived results clearly show the impact of misalignment error, severe atmospheric conditions, type of detection scheme, and number of RF links on the system performance. All the analytical results are validated through simulations.

In this study, the authors carry out the performance analysis of an asymmetric dual hop relay system composed of both radio‐frequency (RF) and free‐space optical (FSO) links. The RF link is subject to generalised η − μ distribution, while the channel for FSO link is modelled as gamma–gamma distribution. The decode‐and‐forward relaying phenomena is used, where the relay decodes the received RF signal from the source and converts it into an optical signal using the sub‐carrier intensity‐modulation (SIM) scheme for transmission over the FSO link. The FSO link is subjected to pointing errors and account for both types of detection techniques, i.e. IM/DD and heterodyne detection. Novel exact closed‐form expressions for the probability density function and cumulative distribution function of the equivalent end‐to‐end signal‐to‐noise ratio of the mixed RF/FSO system in terms of Meijer's G function are derived. Capitalising on these derived channel statistics, they provide the new closed‐form expressions of outage probability and the ergodic channel capacity. They also provide the average bit‐error rate for different binary modulations. Furthermore, the Monte Carlo simulations validate the analytical results.

In this paper, we analyze a free space optical (FSO) cooperative communication system utilizing the Gamma-Gamma fading optical links and a single decode-and-forward (DF) erroneous relay. We adopt the subcarrier intensity modulation (SIM) technique for transmitting the M-ary phase shift keying (M-PSK) modulated optical data over an FSO link. For the considered DF-FSO cooperative system, we derive an optimum maximum-likelihood (ML) decoder and a sub-optimal piecewise linear (PL) decoder in the destination of the system. The proposed PL decoder performs very close to the ML decoder and provides significantly reduced decoding complexity as compared to the optimal decoder. Both the ML and PL decoders incorporate the possibility of erroneous transmission by the DF relay and require the average statistics of the source-relay link for decoding the data of the source in the destination. We also derive the approximate average symbol error rate (SER) of the proposed PL decoder for M-PSK constellation.

The exponentiated Weibull (EW) distribution has been recently proposed for the modelling of free-space optical (FSO) links in the presence of finite sized receiver aperture. In this paper, the performance of FSO communication systems over EW is studied. Specifically, the probability density function (PDF) and cumulative distribution function (CDF) of the instantaneous signal-to-noise ratio (SNR), over EW turbulence fading, are studied. The derived statistics of the SNR is utilized to analyse the performance of an FSO communication system over a generalized communication environment with turbulence induced fading, misalignment errors and path loss. New expression for the outage probability is obtained, and exact expressions for the average bit error rate (BER) are derived for various binary modulation schemes. Finally, the obtained analytical results are verified via Monte Carlo simulations.

In a demodulate-and-forward (DF) based cooperative communication system, erroneous relaying of the data leads to degradation in the performance of the destination receiver. However, a maximum likelihood (ML) decoder in the destination can improve the receiver performance. For achieving a diversity gain, the Alamouti space-time block code (STBC) can be used in the DF based cooperative system in a distributed manner. In this paper, we derive an ML decoder of the distributed Alamouti STBC for the DF based cooperative system with two imperfect relaying nodes. We also consider a DF cooperative communication system in which one out of two relays is in outage. A piece-wise linear (PL) decoder for the DF cooperative system with the distributed Alamouti code and one relay in outage is proposed. The PL decoder provides approximately the same performance as that of the ML decoder with reduced decoding complexity. We derive the pairwise error probability (PEP) of the proposed ML decoder with binary phase-shift keying constellation. An optimized transmit power allocation for the relays is performed by minimizing an upper bound of the PEP. It is shown by simulations that the proposed ML decoder enables the DF protocol based cooperative system to outperform the same rate amplify-and-forward protocol based cooperative system when both systems utilize the distributed Alamouti STBC.

In this paper, we derive a maximum-likelihood (ML) decoder for the demodulate-and-forward (DF) based cooperative communication system using Alamouti space-time block code (STBC) in a distributed manner. We also propose a sub-optimal low complexity piece-wise linear (PL) decoder of the distributed Alamouti code in a DF cooperative system in which one relay out of two relays is in outage. The proposed PL decoder does not lead to any significant performance degradation and performs very close to the proposed ML decoder. Moreover, the proposed ML decoder of the DF cooperative system significantly outperforms an amplify-and-forward (AF) based cooperative system when both systems use the same data rate and distributed Alamouti STBC.

In this paper the performance of coded cooperation has been analyzed for independent flat Nakagami-m fading channels. The outage behavior of the cooperative system has been presented and analyzed with two instantaneous variable (i.e. one with instantaneous Signal to noise ratio (SNR) and other with instantaneous received power). Finally we have presented an analytical approach to calculate the critical cooperation ratio. This gives the value of cooperation ratio ($\alpha$) at which the total outage probability is minimum.