Papers by João Ricardo Mendes
Sensors
Localization is a crucial skill in mobile robotics because the robot needs to make reasonable nav... more Localization is a crucial skill in mobile robotics because the robot needs to make reasonable navigation decisions to complete its mission. Many approaches exist to implement localization, but artificial intelligence can be an interesting alternative to traditional localization techniques based on model calculations. This work proposes a machine learning approach to solve the localization problem in the RobotAtFactory 4.0 competition. The idea is to obtain the relative pose of an onboard camera with respect to fiducial markers (ArUcos) and then estimate the robot pose with machine learning. The approaches were validated in a simulation. Several algorithms were tested, and the best results were obtained by using Random Forest Regressor, with an error on the millimeter scale. The proposed solution presents results as high as the analytical approach for solving the localization problem in the RobotAtFactory 4.0 scenario, with the advantage of not requiring explicit knowledge of the exa...
Communications in computer and information science, 2022
Forests are remote areas with uneven terrain, so it is costly to map the range of signals that en... more Forests are remote areas with uneven terrain, so it is costly to map the range of signals that enable the implementation of systems based on wireless and long-distance communication. Even so, the interest in Internet of Things (IoT) functionalities for forest monitoring systems has increasingly attracted the attention of several researchers. This work demonstrates the development of a platform that uses the GPS technology of mobile devices to map the signals of a LoRaWAN Gateway. Therefore, the proposed system is based on concatenating two messages to optimize the LoRaWAN transmission using the Global Position System (GPS) data from a mobile device. With the proposed approach, it is possible to guarantee the data transmission when finding the ideal places to fix nodes regarding the coverage of LoRaWAN because the Gateway bandwidth will not be fulfilled. The tests indicate that different changes in the relief and large bodies drastically affect the signal provided by the Gateway. This work demonstrates that mapping the Gateway's signal is essential to attach modules in the forest, agriculture zones, or even smart cities.
Cancers
The One Step Nucleic Acid Amplification (OSNA) is being adopted worldwide for sentinel lymph node... more The One Step Nucleic Acid Amplification (OSNA) is being adopted worldwide for sentinel lymph nodes (SLNs) staging in breast cancer (BC). As major disadvantage, OSNA precludes prognostic information based on structural evaluation of SLNs. Our aim is to identify biomarkers related to tumor-microenvironment interplay exploring gene expression data from the OSNA remaining lysate. This study included 32 patients with early stage hormone receptors-positive BC. Remaining OSNA lysates were prepared for targeted RNA-sequencing analysis. Identification of differentially expressed genes (DEGs) was performed by DESeq2 in R and data analysis in STATA. The results show that, in metastatic SLNs, several genes were upregulated: KRT7, VTCN1, CD44, GATA3, ALOX15B, RORC, NECTIN2, LRG1, CD276, FOXM1 and IGF1R. Hierarchical clustering analysis revealed three different clusters. The identified DEGs codify proteins mainly involved in cancer aggressiveness and with impact in immune response. The overexpres...
Background Lymph nodes (LNs) are the main doorway for tumor cell metastases from the primary site... more Background Lymph nodes (LNs) are the main doorway for tumor cell metastases from the primary site and its evaluation is a major prognostic factor. The One Step Nucleic Acid Amplification (OSNA) is being adopted worldwide for sentinel-LNs (SLNs) staging in breast cancer (BC). SLNs´ OSNA lysate may be used for gene expression studies, being the potentially ideal samples to search for new markers related to immune response. Using a targeted gene expression approach, we aim to identify transcriptomic patterns of SLNs immune response and biomarkers that may improve risk stratification and personalized therapy for patients with Luminal A BC. Methods This was an observational, prospective, pilot study that included 32 patients with Luminal A early-stage BC: 16 patients with OSNA negative SLNs and 16 patients with OSNA positive SLNs. After the OSNA assay, rather than being discarded, the remaining OSNA lysates were prepared for target RNA sequencing (RNA-seq) analysis, using the Oncomine™ I...
Optimizing Data Transmission in a Wireless Sensor Network Based on LoRaWAN Protocol
Communications in Computer and Information Science, 2021
For decades, the soft-kill strategy which detects high-speed underwater targets with passive sona... more For decades, the soft-kill strategy which detects high-speed underwater targets with passive sonar and attract the targets using decoys has been used. However, recently, the paradigm to respond to the high-speed underwater targets is shifting to a hard-kill method that directly intercepts the targets. Therefore, an active sonar detection and tracking technique is required to estimate the exact location of the target. Existing detection and tracking techniques using active sonar divide pulses either temporally or along the frequency axis, transmitting pulses in various directions within a pulse repetition interval. This division, however, leads to a reduction in the time-bandwidth product, consequently diminishing detection performance. Therefore, this paper proposes a generalized sinusoidal frequency modulation (GSFM)-based pulsed active sonar (PAS). The proposed PAS-GSFM employs short pulses for a quick update of target estimates, but the orthogonality between pulses of GSFM allows pulses to maximize bandwidth, thus high detection performance can be expected. Two types of simulations were performed to verify the performance of the proposed PAS-GSFM. First, the performance comparison with PAS-linear frequency modulation (LFM), and second, the comparison between the proposed method and continuous active sonar (CAS)-GSFM. Through the evaluation, the superiority of the proposed method over PAS-LFM and competitiveness compared to CAS-GSFM were proved. INDEX TERMS High-speed underwater target, GSFM, pulsed active sonar, target detection and tracking.
For decades, the soft-kill strategy which detects high-speed underwater targets with passive sona... more For decades, the soft-kill strategy which detects high-speed underwater targets with passive sonar and attract the targets using decoys has been used. However, recently, the paradigm to respond to the high-speed underwater targets is shifting to a hard-kill method that directly intercepts the targets. Therefore, an active sonar detection and tracking technique is required to estimate the exact location of the target. Existing detection and tracking techniques using active sonar divide pulses either temporally or along the frequency axis, transmitting pulses in various directions within a pulse repetition interval. This division, however, leads to a reduction in the time-bandwidth product, consequently diminishing detection performance. Therefore, this paper proposes a generalized sinusoidal frequency modulation (GSFM)-based pulsed active sonar (PAS). The proposed PAS-GSFM employs short pulses for a quick update of target estimates, but the orthogonality between pulses of GSFM allows pulses to maximize bandwidth, thus high detection performance can be expected. Two types of simulations were performed to verify the performance of the proposed PAS-GSFM. First, the performance comparison with PAS-linear frequency modulation (LFM), and second, the comparison between the proposed method and continuous active sonar (CAS)-GSFM. Through the evaluation, the superiority of the proposed method over PAS-LFM and competitiveness compared to CAS-GSFM were proved. INDEX TERMS High-speed underwater target, GSFM, pulsed active sonar, target detection and tracking.
For decades, the soft-kill strategy which detects high-speed underwater targets with passive sona... more For decades, the soft-kill strategy which detects high-speed underwater targets with passive sonar and attract the targets using decoys has been used. However, recently, the paradigm to respond to the high-speed underwater targets is shifting to a hard-kill method that directly intercepts the targets. Therefore, an active sonar detection and tracking technique is required to estimate the exact location of the target. Existing detection and tracking techniques using active sonar divide pulses either temporally or along the frequency axis, transmitting pulses in various directions within a pulse repetition interval. This division, however, leads to a reduction in the time-bandwidth product, consequently diminishing detection performance. Therefore, this paper proposes a generalized sinusoidal frequency modulation (GSFM)-based pulsed active sonar (PAS). The proposed PAS-GSFM employs short pulses for a quick update of target estimates, but the orthogonality between pulses of GSFM allows pulses to maximize bandwidth, thus high detection performance can be expected. Two types of simulations were performed to verify the performance of the proposed PAS-GSFM. First, the performance comparison with PAS-linear frequency modulation (LFM), and second, the comparison between the proposed method and continuous active sonar (CAS)-GSFM. Through the evaluation, the superiority of the proposed method over PAS-LFM and competitiveness compared to CAS-GSFM were proved. INDEX TERMS High-speed underwater target, GSFM, pulsed active sonar, target detection and tracking.
For decades, the soft-kill strategy which detects high-speed underwater targets with passive sona... more For decades, the soft-kill strategy which detects high-speed underwater targets with passive sonar and attract the targets using decoys has been used. However, recently, the paradigm to respond to the high-speed underwater targets is shifting to a hard-kill method that directly intercepts the targets. Therefore, an active sonar detection and tracking technique is required to estimate the exact location of the target. Existing detection and tracking techniques using active sonar divide pulses either temporally or along the frequency axis, transmitting pulses in various directions within a pulse repetition interval. This division, however, leads to a reduction in the time-bandwidth product, consequently diminishing detection performance. Therefore, this paper proposes a generalized sinusoidal frequency modulation (GSFM)-based pulsed active sonar (PAS). The proposed PAS-GSFM employs short pulses for a quick update of target estimates, but the orthogonality between pulses of GSFM allows pulses to maximize bandwidth, thus high detection performance can be expected. Two types of simulations were performed to verify the performance of the proposed PAS-GSFM. First, the performance comparison with PAS-linear frequency modulation (LFM), and second, the comparison between the proposed method and continuous active sonar (CAS)-GSFM. Through the evaluation, the superiority of the proposed method over PAS-LFM and competitiveness compared to CAS-GSFM were proved. INDEX TERMS High-speed underwater target, GSFM, pulsed active sonar, target detection and tracking.
Individuals in affiliate marketing programs sign up with companies to promote or sell their produ... more Individuals in affiliate marketing programs sign up with companies to promote or sell their products in independent venues and channels, receiving compensations for their actions. While a component of the e-commerce ecosystem for over a decade, affiliate marketing is increasingly being adopted by companies given its promises of boosting revenue at low investment costs. This work analyzes Clube Hurb, the affiliate marketing program of Hurb.com the largest online travel agency in Brazil. The analysis reveals the fragility of social network growth (very low virality coefficient) along with the strength of social referrals. It also reveals that almost all revenue generated by the program comes from a small set of affiliates, a property that has sustained over time. Indeed, great disparities are characterized by heavy-tailed distributions in statistics concerning both the network and revenue structure. Thus, while most affiliates play no effective role, a few are instrumental in keeping the program profitable. Such findings indicate that traditional average-based performance metrics can be flawed when assessing the success of such programs.
The teleoperation of UAVs often demands extensive training, since even well trained pilots are pr... more The teleoperation of UAVs often demands extensive training, since even well trained pilots are prone to mistakes, resulting frequently in collisions of the vehicle with obstacles. This paper presents a method to assist the teleoperation of a quadrotor using an obstacle avoidance approach. The target scenario is unknown, unstructured, and GPS-denied. A short-term rough map of the nearby environment is constructed using sonar sensors. This map is constructed using FastSLAM to allow tracking of the vehicle position with respect to the map. A danger classification method is then applied to choose the appropriate action for each particular, and potentially dangerous, situation. A simple active perception routine is used to orient one of the sensors to an unknown area, in case the UAV is ordered to move towards an unmapped area. Real world results are presented allowing a preliminary validation of the proposed methods.
2012 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR), 2012
Unmanned Aerial Vehicles (UAV) provide many advantages in Search and Rescue (SaR) scenarios, such... more Unmanned Aerial Vehicles (UAV) provide many advantages in Search and Rescue (SaR) scenarios, such as the capacity for remote inspection over areas that are difficult to reach by ground vehicles. Moreover, it can carry small payloads, such as first aid equipment, over large distances. However, the teleoperation of UAVs often demands extensive training, since even well trained pilots are prone to mistakes, resulting frequently in collisions of the vehicle with obstacles. This paper presents a method to assist the teleoperation of a quadrotor using an obstacle avoidance approach. The target scenario is SaR operation in unknown, unstructured, GPSdenied environments, such as warehouses or other buildings. A shortterm rough map of the nearby environment is constructed using sonar sensors. This map is constructed using FastSLAM to allow tracking of the vehicle position with respect to the map. The map is then used to (1) override operator commands that may lead to a collision, and (2) perform evasive maneuvers whenever collision is imminent. A simple active perception routine is used to orient one of the sensors to an unknown area, in case the UAV is ordered to move towards an unmapped area. Experimental results using the USARsim simulator are presented. Further testing was conducted in a real quadcopter, allowing a preliminary validation of the proposed methods.
Universidade Presbiteriana Mackenzie, Jan 24, 2007
Anais do XXV Simpósio Brasileiro de Telecomunicações, 2007
Based on concurrent algorithms and on the convex combination of one slow and one fast CMA (Consta... more Based on concurrent algorithms and on the convex combination of one slow and one fast CMA (Constant Modulus Algorithm), we propose a convex combination of two blind equalizers adapted respectively by CMA and the modified SDD (Soft Decision-Directed) algorithm for recovering of QAM (Quadrature Amplitude Modulation) signals. For high signal-to-noise ratio, the performance of the proposed scheme is, in the worst case, as good as that of the best of its components, being better than both of them in some situations. Since the considered algorithms are derived from different criteria, the mixing parameter is updated for minimizing the decision-directed cost function.
2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2011
We propose blind equalization algorithms that perform similarly to supervised ones, independently... more We propose blind equalization algorithms that perform similarly to supervised ones, independently of the QAM order. They converge approximately to the Wiener solution, which generally provides a relatively low misadjustment. Besides presenting strategies to speed up their convergences, we provide sufficient conditions for the stability of the symbol-based decision algorithm, which is an extension of the decision-directed algorithm. Their behaviors are illustrated through simulation results.
A region-based algorithm for blind equalization of QAM signals
2009 IEEE/SP 15th Workshop on Statistical Signal Processing, 2009
We propose a region-based multimodulus algorithm for blind equalization of high-order quadrature ... more We propose a region-based multimodulus algorithm for blind equalization of high-order quadrature amplitude modulation (QAM) signals. It treats nonconstant modulus constellations as constant modulus ones, converging approximately to the Wiener solution. To avoid divergence, it rejects non-consistent estimates of the transmitted signal. When compared to existing blind multimodulus-type algorithms for equalization of QAM signals, it exhibits considerably lower misadjustment, faster
Accelerating the convergence of a decision-based algorithm for blind equalization of QAM signals
2012 3rd International Workshop on Cognitive Information Processing (CIP), 2012
ABSTRACT Recently, we proposed a decision-based algorithm for blind equalization that performs si... more ABSTRACT Recently, we proposed a decision-based algorithm for blind equalization that performs similarly to a supervised algorithm in steady-state, independently of the QAM order. Using this algorithm, the usual switching to the decision-directed mode may be avoided. Additionally, a low-cost scheme based on the neighborhood of the estimated symbol was used to improve its convergence rate. In this paper, we interpret this scheme from a probability density function fitting perspective. When the uncertainty on the estimated symbol is high, simulations show that the neighborhood plays an important role for speeding up the convergence. On the other hand, to ensure a low mean-square error, the neighborhood must be disregarded in steady-state.
Signal Processing, 2012
It is well-known that constant-modulus-based algorithms present a large mean-square error for hig... more It is well-known that constant-modulus-based algorithms present a large mean-square error for high-order quadrature amplitude modulation (QAM) signals, which may damage the switching to decision-directedbased algorithms. In this paper, we introduce a regional multimodulus algorithm for blind equalization of QAM signals that performs similarly to the supervised normalized least-mean-squares (NLMS) algorithm, independently of the QAM order. We find a theoretical relation between the coefficient vector of the proposed algorithm and the Wiener solution and also provide theoretical models for the steady-state excess mean-square error in a nonstationary environment. The proposed algorithm in conjunction with strategies to speed up its convergence and to avoid divergence can bypass the switching mechanism between the blind mode and the decision-directed mode.
Space-time blind decision feedback equalizer
2006 International Telecommunications Symposium, 2006
We address the problem of signal separation using space-time blind decision feedback equalizer. A... more We address the problem of signal separation using space-time blind decision feedback equalizer. Assuming correct decisions and absence of noise, the perfect equalization conditions are obtained. We present an extension of the blind algorithm which avoids degenerated solutions in the single-input single output case. The proposed algorithm jointly adapts the feedforward and feedback filters of DFE, avoids degenerated solutions, and has capability of simultaneously recovering all sources.
Electronics Letters, 2012
Proposed is a symbol-based decision-directed algorithm for blind equalisation of quadrature ampli... more Proposed is a symbol-based decision-directed algorithm for blind equalisation of quadrature amplitude modulation (QAM) signals using a decision feedback scheme. Independently of QAM order, it presents: (i) an error equal to zero when the equaliser output coincides with the transmitted signal; (ii) simultaneous recovery of the modulus and phase of the signal; (iii) a misadjustment close to that of the normalised least-mean squares algorithm; (iv) fast convergence; and (v) the avoidance of degenerative solutions. Additionally, its stability is ensured when the step-size is properly chosen.
Uploads
Papers by João Ricardo Mendes