Probability Of False Detection Research Articles

A non-zero Doppler amplitude is not enough: revisiting the putative radial velocity detection of sub-Venus exoplanet L 98-59b

ABSTRACT L 98-59b is a transiting exoplanet with radius ${0.85^{+0.061}_{-0.047}}$R⊕ and orbital period ${2.2531136^{+1.2e-06}_{-1.5e-06}}$ d. In 2021, a Doppler detection of L 98-59b was announced, with radial velocity (RV) semi-amplitude $K_{\rm b}=0.46^{+0.20}_{-0.17}$$\textrm {m}\, \textrm {s}^{-1}$ inferred from 164 High Accuracy Radial velocity Planet Searcher and 66 Échelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observation RVs. This translated into a mass $M_{\rm b}=0.40^{+0.16}_{-0.15}$M⊕, or half the mass of Venus, making L 98-59b then (by far) the lowest mass planet with a putative RV detection. To illuminate the difficulties in securely detecting <1 $\textrm {m}\, \textrm {s}^{-1}$ signals, we argue here that the published RV detection of L 98-59b is not supported by sufficient statistical evidence. We show, under the published modelling assumptions, that Bayesian model comparison (BMC) suggests an $\sim 50~{{\ \rm per\ cent}}$ false-detection probability. We also often infer ∼40 $\textrm {cm}\, \textrm {s}^{-1}$ semi-amplitudes for Keplerians with periods not corresponding to any known planet; importantly, though, BMC rejects all these ‘detections’. By implementing a more sophisticated stellar activity model and more realistic parameter priors, we infer a mass Mb = 0.47 ± 0.14M⊕ from semi-amplitude Kb = 0.56 ± 0.16 $\textrm {m}\, \textrm {s}^{-1}$, with BMC now indicating substantial yet still not decisive evidence for the detection. Definitive detection and characterization may require many additional RVs, alongside careful modelling and stringent statistical analysis. Our case study has implications for current extreme-precision RV campaigns and the longer term quest to discover the first Earth-twins.

High Sensitivity and High Throughput Magnetic Flow CMOS Cytometers with 2D Oscillator Array and Inter-Sensor Spectrogram Cross-correlation.

In the paper, we present an integrated flow cytometer with a 2D array of magnetic sensors based on dual-frequency oscillators in a 65-nm CMOS process, with the chip packaged with microfluidic controls. The sensor architecture and the presented array signal processing allows uninhibited flow of the sample for high throughput without the need for hydrodynamic focusing to a single sensor. To overcome the challenge of sensitivity and specificity that comes as a trade off with high throughout, we perform two levels of signal processing. First, utilizing the fact that a magnetically tagged cell is expected to excite sequentially an array of sensors in a time-delayed fashion, we perform inter-site cross-correlation of the sensor spectrograms that allows us to suppress the probability of false detection drastically, allowing theoretical sensitivity reaching towards sub-ppM levels that is needed for rare cell or circulating tumor cell detection. In addition, we implement two distinct methods to suppress correlated low frequency drifts of singular sensors-one with an on-chip sensor reference and one that utilizes the frequency dependence of the susceptibility of super-paramagnetic magnetic beads that we deploy as tags. We demonstrate these techniques on a 7×7 sensor array in 65 nm CMOS technology packaged with microfluidics with magnetically tagged dielectric particles and cultu lymphoma cancer cells.

A locational false data injection attack detection method in smart grid based on adversarial variational autoencoders

Stealthy FDIA (False Data Injection Attack) is a serious cyber threat that can modify state estimation of smart grid through maliciously altering the measurement data, but can’t be detected by traditional bad data detection system in smart grid. There exist two weakpoints for numerous deep neural networks (DNNs) based data-driven schemes against FDIA. First, they mainly focus on detecting the presence of FDIA, but fail to localize the specific bus/nodes affected. Second, their performance is not sufficiently desirable under small attack, i.e., anomalies caused by the attack closely resemble normal data. To address the above issues, this paper proposes an effective locational FDIA detection framework based on data reconstruction, AT-GVAE, which seamlessly integrates the variational autoencoders (VAEs) and generative adversarial network paradigm. Specifically, our contributions are threefold. First, the proposed AT-GVAE framework is novelly composed of two main modules: the generative VAE_G and discriminative VAE_D that both play dual roles: reconstruct data from jointly learning distributions of data and latent feature space, and meanwhile play Minmax game with adversarial way. Second, multiple-layer gated recurrent units (GRUs) are utilized as the basic structure of the encoder and decoders in both VAEs, to characterize the temporal correlations of measurement data sequence. Additionally, the self-attention mechanism is used to enhance the expressive ability of GRU based VAEs. Then, the anomaly score for each busbar in the smart grid is determined by comparing the residual between the observed measurement and the outputs of VAE_G and VAE_D, enabling the localization of FDIA. Finally, thorough experiments on multiple power systems with series data of total 3456 measurements demonstrate that, in terms of multiple typical metrics, including false alarm rate vs. detection probability, AUC-ROC, and AUC-PR, our proposed framework outperforms the state-of-the-art GRU, VAE and adversarial training based FDIA detection schemes.

GLRT‐based compressive subspace detectors in single‐frequency multistatic passive radar systems

AbstractThe authors study the problem of compressive target detection in a single‐frequency network (SFN)‐based multistatic passive radar system (MS‐PRS) consisting of multiple illuminators of opportunity (IOs) and one receiver. Firstly, a generalised likelihood ratio test (GLRT)‐based SFN‐based compressive subspace detector (SFN‐CSD) is derived by exploiting the sparsity of the target echoes for the case of known noise variance. When the noise variance is unknown, an SFN‐based unknown‐noise (UN) compressive subspace detector is proposed, referred to as the SFN‐UNCSD. Moreover, closed‐form expressions of the probability of false alarm and detection of the proposed detectors are deriived. It is proved that the SNF‐UNCSD has a constant false alarm rate (CFAR) property. Finally, numerical simulations are conducted to verify the theoretical analysis and illustrate the performance of the proposed detector relative to several benchmark detectors.

Joint detection algorithm for multiple cognitive users in spectrum sensing

Spectrum sensing technology is a crucial aspect of modern communication technology, serving as one of the essential techniques for efficiently utilizing scarce information resources in tight frequency bands. This paper first introduces three common logical circuit decision criteria in hard decisions and analyzes their decision rigor. Building upon hard decisions, the paper further introduces a method for multi-user spectrum sensing based on soft decisions. Then the paper simulates the false alarm probability and detection probability curves corresponding to the three criteria. The simulated results of multi-user collaborative sensing indicate that the simulation process significantly reduces false alarm probability and enhances detection probability. This approach effectively detects spectrum resources unoccupied during idle periods, leveraging the concept of time-division multiplexing and rationalizing the redistribution of information resources. The entire computation process relies on the calculation principles of power spectral density in communication theory, involving threshold decision detection for noise power and the sum of noise and signal power. It provides a secondary decision detection, reflecting the perceptual decision performance of logical detection methods with relative accuracy.

Tag-Based PHY-Layer Authentication for RIS-Assisted Communication Systems

This paper proposes a tag-based approach for physical (PHY)-layer authentication in a reconfigurable intelligent surface (RIS) communication system. We first extract the intrinsic PHY-layer features of RIS communication systems in terms of channel gain and background noise, and then apply these PHY-layer features, a random signal as well as the private key of the transmitter to construct a robust cover tag signal against the impersonation attack. We adopt an asymmetric cryptography technique to encrypt tagged signals and to resist against unauthorized detection and tampering attacks during the transmission process. The receiver then applies the maximum a-posteriori (MAP) ratio test to conduct authentication based on the received tag signal, a reference tag signal transmitted in training phase and the knowledge of distributions of the channel gain, background noise and the random signal. We also provide security analysis to demonstrate how the proposed scheme can resist unauthorized detection, tampering attacks, etc. With the help of tools of the MAP ratio test, maximum likelihood estimation, we further analyze the distribution of the test statistics and derive analytical models for the false alarm and detection probabilities. Finally, extensive simulations are conducted to verify the theoretical results and to illustrate the performance of the proposed scheme.

Real-Time Traffic Flow Statistics Based on Dual-Granularity Classification

Article Real-Time Traffic Flow Statistics Based on Dual-Granularity Classification Yanchao Bi, Yuyan Yin, Xinfeng Liu *, Xiushan Nie, Chenxi Zou, and Junbiao Du 1 School of Computer Science and Technology, Shandong Jianzhu University, Jinan 250101, China * Correspondence: liuxinfeng18@sdjzu.edu.cn Received: 17 June 2023 Accepted: 13 September 2023 Published: 26 September 2023 Abstract: Traffic detection devices can cause accuracy degradation over time. Considering problems such as time-consuming and laborious manual statistics, high misdetection probabilities, and model tracking failures, there is an urgent need to develop a deep learning model (which can stably achieve detection accuracy over 90%) to evaluate whether the device accuracy still satisfies the requirements or not. In this study, based on dual-granularity classification, a real-time traffic flow statistics method is proposed to address the above problems. The method is divided into two stages. The first stage uses YOLOv5 to acquire all the motorized and non-motorized vehicles appearing in the scene. The second stage uses EfficientNet to acquire the motorized vehicles obtained in the previous stage and classify such vehicles into six categories. Through this dual-granularity classification, the considered problem is simplified and the probability of false detection is reduced significantly. To correlate the front and back frames of the video, vehicle tracking is implemented using DeepSORT, and vehicle re-identification is implemented in conjunction with the ResNet50 model to improve the tracking accuracy. The experimental results show that the method used in this study solves the problems of misdetection and tracking effectively. Moreover, the proposed method achieves 98.7% real-time statistical accuracy by combining the two-lane counting method.

Evaluation of Seasonal Differences among Three NOAA Climate Data Records of Precipitation

Abstract Three satellite precipitation datasets—CMORPH, PERSIANN-CDR, and GPCP—from the NOAA/Climate Data Record program were evaluated in their ability to capture seasonal differences in precipitation for the period 2007–18 over the conterminous United States. Data from the in situ U.S. Climate Reference Network (USCRN) provided reference precipitation measurements and collocated atmospheric conditions (temperature) at the daily scale. Satellite precipitation products’ (SPP) performance with respect to cold season precipitation was compared to warm season and full-year analysis for benchmarking purposes. Considering an ensemble of typical performance metrics including accuracy, false alarm ratio, probability of detection, probability of false detection, and the Kling–Gupta efficiency (KGE) that combines correlation, bias, and variability, we found that the three SPPs displayed better performances during the warm season than during the cold season. Among the three datasets, CMORPH displayed better performance—smaller bias, higher correlation, and a better KGE score—than the two other SPPs on an annual basis and during the warm season. During the cold season, CMORPH showed the worst performance at higher latitudes over areas experiencing recurring snow or frozen and mixed precipitation. CMORPH’s performances were further degraded compared to PERSIANN-CDR and GPCP when considering freezing temperatures (T < 0°C) due to the inability to microwave sensors to retrieve precipitation over snow-covered surface. However, for cold rainfall events detected simultaneously by the satellite and the USCRN stations (i.e., conditional case), CMORPH performance noticeably improved but remained inferior to the two other datasets. The quantification of seasonal precipitation errors and biases for three satellite precipitation datasets presented in this work provides an objective basis for the improvement of rainfall retrieval algorithms of the next generation of satellite precipitation products.

Adaptive Channel Detection for Full-Duplex IAB Systems on Unlicensed Bands

The integration of integrated access and backhaul (IAB) and full-duplex communications on unlicensed bands can improve the spectrum efficiency and mitigate the spectrum resource shortage. To guarantee fair coexistence on unlicensed spectrum, the listen-before-talk mechanism needs to be executed at the IAB nodes, where the clear channel assessment (CCA) is applied before transmission. However, the residual self-interference (RSI) brought by the full-duplex communication may degrade the CCA detection performance tremendously. Therefore, in this paper we propose an adaptive channel detection scheme for the full-duplex IAB network to minimize the impact of RSI on the CCA detection, which can enhance the spectrum efficiency. In particular, the closed-form expressions of false alarm probability and detection probability are derived. Accordingly, the energy detection (ED) threshold applied in CCA is adjusted adaptively to improve the system throughput. Then, a novel timing sequence with respect to the self-interference cancellation and CCA detection is proposed. The corresponding optimal initial threshold is also derived to realize the adaptive ED threshold. Furthermore, an enhanced request to send/clear to send mechanism is proposed to solve the hidden terminal problem. Simulation results demonstrate that the proposed schemes can effectively reduce the impact of RSI on the CCA detection accuracy and significantly improve the system throughput.

Adaptive Jamming Attack Detection Under Noise Uncertainty in mmWave Massive MIMO Systems

This paper addresses jamming attack detection issue in a millimeter Wave (mmWave) massive MIMO system under noise uncertainty. Specifically, we apply the generalized likelihood ratio test (GLRT) to develop a one-step GLRT scheme for detecting jamming attack under the homogeneous and partially homogeneous noise environments, and exploit training data to estimate the unknown noise statistical information and replace it with resulting estimation in deriving GLRT procedure to obtain adaptive GLRT detection scheme. We also design a two-step GLRT scheme, where we first assume the unknown noise statistical information is known, and derive GLRT based on test data, and then replace it by the sample covariance matrix based on training data only to achieve a fully adaptive jamming attack detector. With the help of statistical signal subspace analysis and composite hypothesis testing theories, we further examine the statistical distributions of the jamming attack detection schemes and present the closed-form expressions of false alarm and detection probabilities for the proposed schemes under different noise environments. Finally, we implement extensive simulations to validate the theoretical results and evaluate the detection efficiency under various parameters.

ED-Based Spectrum Sensing for the Satellite Communication Networks Using Phased-Array Antennas

Spectrum sensing is of crucial importance in the satellite communication networks to address the increasingly severe spectrum scarcity problem. However, most of the spectrum sensing schemes for detecting satellite signal use the traditional reflector antennas and have poor performance, especially under the low signal-to-noise ratio (SNR). To tackle this issue, an energy detection (ED)-based spectrum sensing scheme is proposed by using the phased-array antennas (PAAs) for the satellite communication networks. Moreover, the closed-form expressions for the false alarm probability and detection probability considering the visibility of satellite-to-ground links are derived. Simulation results verify our theoretical analysis and demonstrate the superiority of our proposed scheme compared with the schemes that do not consider the spatial condition.

Sequential detection of transient signal by moving likelihood ratio statistic in an exponential family

ABSTRACT We first consider the sequential detection of transient signals by generalizing the moving average chart to exponential family and study the false detection probability (FDP) and power of detection (POD) in the steady state. Then windowed adjusted signed (or modified directed) likelihood ratio chart is studied by treating it as approximate normal random variable. In the multi-parameter exponential family, the detection of the transient change of one of the canonical parameters or a function of canonical parameters is considered by using the generalized adjusted signed likelihood ratio chart. Comparisons with window restricted CUSUM and Shiryayev-Roberts (S-R) procedures show that the generalized signed likelihood ratio chart performs quite well. Several important examples including the mean or variance change under normal model and a real-time example are used for illustration.

Naïve Bayesian Classification Based Glioma Brain Tumor Segmentation Using Grey Level Co-occurrence Matrix Method

Brain tumors vary widely in size and form, making detection and diagnosis difficult. This study's main aim is to identify abnormal brain images., classify them from normal brain images, and then segment the tumor areas from the categorised brain images. In this study, we offer a technique based on the Nave Bayesian classification approach that can efficiently identify and segment brain tumors. Noises are identified and filtered out during the preprocessing phase of tumor identification. After preprocessing the brain image, GLCM and probabilistic properties are extracted. Naive Bayesian classifier is then used to train and label the retrieved features. When the tumors in a brain picture have been categorised, the watershed segmentation approach is used to isolate the tumors. This paper's brain pictures are from the BRATS 2015 data collection. The suggested approach has a classification rate of 99.2% for MR pictures of normal brain tissue and a rate of 97.3% for MR images of aberrant Glioma brain tissue. In this study, we provide a strategy for detecting and segmenting tumors that has a 97.54% Probability of Detection (POD), a 92.18% Probability of False Detection (POFD), a 98.17% Critical Success Index (CSI), and a 98.55% Percentage of Corrects (PC). The recommended Glioma brain tumour detection technique outperforms existing state-of-the-art approaches in POD, POFD, CSI, and PC because it can identify tumour locations in abnormal brain images.

Pull & Push: Leveraging Differential Knowledge Distillation for Efficient Unsupervised Anomaly Detection and Localization

Recently, much attention has been paid to segmenting subtle unknown defect regions by knowledge distillation in an unsupervised setting. Most previous studies concentrated on guiding the student network to learn the same representations on the normality, neglecting the different behaviors of the abnormality. This leads to a high probability of false detection of subtle defects. To address such an issue, we propose to push representations on abnormal areas of the teacher and student network as far as possible while pulling representations on normal areas as close as possible. Based on this idea, we design an efficient teacher-student model for anomaly detection and localization, which maximizes pixel-wise discrepancies for anomalous regions approximated by data augmentation and simultaneously minimizes discrepancies for pixel-wise normal regions between these two networks. The explicit differential knowledge distillation enlarges the margin between normal representations and abnormal ones in favour of discriminating them. Then, the appropriate small student network is not only efficient, but more importantly, helps inhibit the generalization ability of anomalous patterns when learning normal patterns, facilitating the precise decision boundary. The experimental results on the MVTec AD, Fashion-MNIST, and CIFAR-10 datasets demonstrate that our proposed method achieves better performance than current state-of-the-art (SOTA) approaches. Especially, For the MVTec AD dataset with high resolution images, we achieve 98.1 AUROC% and 93.6 AUPRO% in anomaly localization, outperforming knowledge distillation based SOTA methods by 1.1 AUROC% and 1.5 AUPRO% with a lightweight model.

Performance Analysis of Uncoordinated Interference Mitigation for Automotive Radar

As a key sensor for Advanced Driving Assistance System (ADAS), millimeter automotive radar has been a promising candidate for fulfilling tasks including adaptive cruise control and collision avoidance. However, the widely deployment of millimeter automotive radars may cause serious mutual interference among vehicles, thus degrading radar ranging performance severely. In this article, we analyze the mutual interference among multiple Frequency Modulated Continuous Wave (FMCW) radars. On one hand, we model the interference precisely by employing Matern Hard-Core Process (MHCP) model to characterize the distribution of vehicle nodes in practical bidirectional two-lane and multi-lane scenarios. Besides, the interference is analyzed in terms of the channel fading, the directional antenna pattern and the fluctuation of the target Radar Cross-Section (RCS) in two-lane and multi-lane scenarios. Besides, we analyze the reflected interference in detail. On the other hand, we evaluate the interference mitigation performance of the Random Frequency Division Multiplexing (RFDM) and Frequency Hopping (FH) approaches in terms of the probability of false detection and miss detection, effective detectable density and maximum number of interference-free radar. Finally, a novel AFH-PM mitigation approach is proposed to further improve the interference mitigation performance, which combines the adaptive FH technology with the binary phase modulation. Simulation results verify the proposed framework for interference analysis by employing Monte Carlo method, and the performance improvement of RFDM, FH and AFH-PM is 6.7 dB, 7.6 dB and 8.2 dB, respectively.

Reviewing the Status of Droughts, Early Warning Systems and Climate Services in South India: Experiences Learned

Drought is one of the most challenging disasters that impact the natural and cultural ecosystems across the world, especially in the climate dependent sectors of arid and semi-arid areas. The aim of this article is to share the experiences gained and enhance the readers’ awareness on the status of drought and process of the early warning systems (EWS) in south India. Drought status of three agroecologically different states is included in this article, such as Kerala, Tamil Nadu and Telangana. As far as Tamil Nadu is concerned, Karur, Thuthukudi, Krishnagiri, Namakkal, Trichy and Thirunelveli districts are water scarce compared to other districts in the state. The districts such as Wayanad, Thiruvananthapuram, Idukki and Palakkad in Kerala have received lesser rainfall compared to the other parts of the state during the period 1981 to 2019. In Telangana, the mandals such as Nagarkurnool, Jogulamba-Gadwal, Wanaparthy, Mahabubnagar Nalgonda and Yedadri are frequently hit by dry spells and droughts. As a case study, weather early warning dissemination, carried out at Parambikulam Aliyar basin, Coimbatore, Tamil Nadu, during Khariff and Rabi seasons, using IMDs medium and extended range forecast is also elaborated in particular in the article. As far as the accuracy of forecast is concerned, probability of false detection (false alarm rate) was found to be 0.81 for Khariff and 0.30 for Rabi season, indicating the need for better performance in the accuracy of dry spell early warning, disaster preparedness and response. In-spite of this, access to early warning has supported the farmers during harvest and land preparation with a utility score of 72% and 59%, respectively. In Parambikulam Aliyar basin, remote sensing products such as MODIS-NDVI, NDWI and TWI was also used to identify the real-time progression of monthly vegetative condition for Kharif and Rabi seasons. NDVI values were used to monitor the district level vegetation condition and compared it with the drought year 2016, the difference in area under barren land was 76% less during Khariff, 2021 and 44% during Rabi, 2021.This study is a compilation of lessons learned from different states and the existing knowledge and practice in early warnings, and recommends the need for a holistic approach in drought and dry spell monitoring along with better accuracy and dissemination to minimize climate-related shocks in agriculture.

Evaluating the IMERG precipitation satellite product to derive intensity-duration-frequency curves in Colombia

This article explores the potentialities of the IMERG V06B FINAL product for estimating Intensity-Duration-Frequency curves in Colombia, using the in-situ data available for 110 rain gauges. From observed data for 76 of these stations, we validated the satellite IMERG precipitation data for the period 2001-2019, at daily, monthly, and annual resolutions. For 60 stations, better results were obtained for the monthly time aggregation, followed by the yearly and daily scales, suggesting that seasonality is the main rainfall characteristic captured by the product. Concerning the occurrence of daily precipitation, results indicate that both the probability of detection and the probability of false detection are high. In general terms, the comparison between intensities from existing IDF curves and those derived from IMERG showed underestimations of the rainfall intensities for the short durations studied (0.5 and 1 h), with mean relative errors in the range [-69%,+56%], and overestimations for the large durations of 2 and 6 h, with mean relative errors in the range [-61%,+171%]. Results also suggest that the IMERG product at this moment is not able to capture the daily rainfall distribution in most of the stations. Nevertheless, for almost 20% of the rain gauges, located mainly in the Amazon, Orinoco, and Pacific Regions, the analysis showed that the maximum intensities derived from IMERG are within +/-25% relative error, compared with the ones calculated using the traditional approach.

Objective Bayesian Approach for Binary Hypothesis Testing of Multivariate Gaussian Observations

This paper proposes an objective Bayesian detector for the binary hypothesis testing problem in which the observation vectors are IID and follow multivariate Gaussian distribution under both the null and alternative hypotheses. The mean vector and covariance matrix under each hypothesis are not known, they also do not have special structure other than the covariance matrices are partially ordered in the positive definite cone with their difference positive semi-definite. An example of such a problem is the detection of Gaussian random signal in Gaussian noise. Non-informative priors are imposed on the unknown parameters for computing the marginals and evaluating the Bayes factor test statistic, where the prior for the covariance matrix is applied to the Cholesky factorization of the precision matrix. For the null hypothesis, we propose conjugate priors for the unknowns. For the alternative hypothesis, we propose uniform prior on the mean vector, and a class of objective priors that encompasses the Jeffreys, Independence Jeffreys, Geisser and Cornfield, Haar Measure and Reference priors for the precision matrix. The proposed priors enable closed-form expressions for the marginals and lead to an objective Bayes factor having the posteriors mainly governed by the data. The proposed detector is analyzed for justifying the priors used, deriving the theoretical moments, and approximating the false alarm and detection probability densities by the shifted Gamma distributions. The developed detector exhibits improvement in detection performance over the energy detector (ED) and the generalized likelihood ratio test (GLRT).

K-Test: An Analytical Model for Measurement-Based Server Selection Based on Response Time

Different paradigms employ server selection to establish a fast connection to end users, thereby reducing the delay associated with exchanging data between mobile users and servers. To efficiently infer the statistics of the server response, we propose an analytical model for server selection based on <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$K$</tex-math></inline-formula> measured response times, namely, the K-Test method. First, the probability that the user device correctly selects the best server and the corresponding time taken in K-Test method are investigated. Then, we study the server switch when connection failure is encountered. The proposed analytical model for K-Test method considers the distribution of the response time of candidate servers and derives three metrics, namely the probability of selecting the best server, the expected latency of the server selection process, and the false connection failure detection probability. The analytical model is validated by the simulation. Numerical results show that both the value of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$K$</tex-math></inline-formula> and the number of candidate servers significantly affect the selection probability. Also, it needs more test rounds to correctly find the best server when users experience a variation in the network response times.

MFF-Net: Multimodal Feature Fusion Network for 3D Object Detection

In complex traffic environment scenarios, it is very important for autonomous vehicles to accurately perceive the dynamic information of other vehicles around the vehicle in advance. The accuracy of 3D object detection will be affected by problems such as illumination changes, object occlusion, and object detection distance. To this purpose, we face these challenges by proposing a multimodal feature fusion network for 3D object detection (MFF-Net). In this research, this paper first uses the spatial transformation projection algorithm to map the image features into the feature space, so that the image features are in the same spatial dimension when fused with the point cloud features. Then, feature channel weighting is performed using an adaptive expression augmentation fusion network to enhance important network features, suppress useless features, and increase the directionality of the network to features. Finally, this paper increases the probability of false detection and missed detection in the non-maximum suppression algorithm by increasing the one-dimensional threshold. So far, this paper has constructed a complete 3D target detection network based on multimodal feature fusion. The experimental results show that the proposed achieves an average accuracy of 82.60% on the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) dataset, outperforming previous state-of-the-art multimodal fusion networks. In Easy, Moderate, and hard evaluation indicators, the accuracy rate of this paper reaches 90.96%, 81.46%, and 75.39%. This shows that the MFF-Net network has good performance in 3D object detection.