References of "Alves Martins, Wallace 50034845"
     in
Bookmark and Share    
Full Text
Peer Reviewed
See detailTowards the Application of Neuromorphic Computing to Satellite Communications
Ortiz Gomez, Flor de Guadalupe UL; Lagunas, Eva UL; Alves Martins, Wallace UL et al

in Towards the Application of Neuromorphic Computing to Satellite Communications (2022, October)

Artificial intelligence (AI) has recently received significant attention as a key enabler for future 5G-and-beyond terrestrial wireless networks. The applications of AI to satellite communications is also ... [more ▼]

Artificial intelligence (AI) has recently received significant attention as a key enabler for future 5G-and-beyond terrestrial wireless networks. The applications of AI to satellite communications is also gaining momentum to realize a more autonomous operation with reduced requirements in terms of human intervention. The adoption of AI for satellite communications will set new requirements on computing processors, which will need to support large workloads as efficiently as possible under harsh environmental conditions. In this context, neuromorphic processing (NP) is emerging as a bio-inspired solution to address pattern recognition tasks involving multiple, possibly unstructured, temporal signals and/or requiring continual learning. The key merits of the technology are energy efficiency and capacity for on-device adaptation. In this paper, we highlight potential use cases and applications of NP to satellite communications. We also explore major technical challenges for the implementation of space-based NP focusing on the available NP chipsets. [less ▲]

Detailed reference viewed: 26 (1 UL)
Full Text
Peer Reviewed
See detailMatching Traffic Demand in GEO Multibeam Satellites: The Joint Use of Dynamic Beamforming and Precoding Under Practical Constraints
Chaker, Haythem UL; Chougrani, Houcine UL; Alves Martins, Wallace UL et al

in IEEE Transactions on Broadcasting (2022)

To adjust for the non-uniform spatiotemporal nature of traffic patterns, next-generation high throughput satellite (HTS) systems can benefit from recent technological advancements in the space-segment in ... [more ▼]

To adjust for the non-uniform spatiotemporal nature of traffic patterns, next-generation high throughput satellite (HTS) systems can benefit from recent technological advancements in the space-segment in order to dynamically design traffic-adaptive beam layout plans (ABLPs). In this work, we propose a framework for dynamic beamforming (DBF) optimization and adaptation in dynamic environments. Given realistic traffic patterns and a limited power budget, we propose a feasible DBF operation for a geostationary multibeam HTS network. The goal is to minimize the mismatch between the traffic demand and the offered capacity under practical constraints. These constraints are dictated by the traffic-aware design requirements, the on-board antenna system limitations, and the signaling considerations in the K-band. Noting that the ABLP is agnostic about the inherent inter-beam interference (IBI), we construct an interference simulation environment using irregularly shaped beams for a large-scale multibeam HTS system. To cope with IBI, the combination of on-board DBF and on-ground precoding is considered. For precoded and non-precoded HTS configurations, the proposed design shows better traffic-matching capabilities in comparison to a regular beam layout plan. Lastly, we provide trade-off analyses between system-level key performance indicators for different realistic non-uniform traffic patterns. [less ▲]

Detailed reference viewed: 53 (4 UL)
Full Text
Peer Reviewed
See detailMaximizing the Number of Served Users in a Smart City using Reconfigurable Intelligent Surfaces
Zivuku, Progress UL; Kisseleff, Steven UL; Nguyen, van Dinh UL et al

in Proceedings of IEEE Wireless Communications and Networking Conference (WCNC) (2022, April 10)

Detailed reference viewed: 38 (4 UL)
Full Text
Peer Reviewed
See detailKernel Regression over Graphs using Random Fourier Features
Elias, Vitor R.M.; Gogineni, Vinay C.; Alves Martins, Wallace UL et al

in IEEE Transactions on Signal Processing (2022)

This paper proposes efficient batch-based and online strategies for kernel regression over graphs (KRG). The proposed algorithms do not require the input signal to be a graph signal, whereas the target ... [more ▼]

This paper proposes efficient batch-based and online strategies for kernel regression over graphs (KRG). The proposed algorithms do not require the input signal to be a graph signal, whereas the target signal is defined over the graph. We first use random Fourier features (RFF) to tackle the complexity issues associated with kernel methods employed in the conventional KRG. For batch-based approaches, we also propose an implementation that reduces complexity by avoiding the inversion of large matrices. Then, we derive two distinct online strategies using RFF, namely, the mini-batch gradient KRG (MGKRG) and the recursive least squares KRG (RLSKRG). The stochastic gradient KRG (SGKRG) is introduced as a particular case of the MGKRG. The MGKRG and the SGKRG are low-complexity algorithms that employ stochastic gradient approximations in the regression-parameter update. The RLSKRG is a recursive implementation of the RFF-based batch KRG. A detailed stability analysis is provided for the proposed online algorithms, including convergence conditions in both mean and mean-squared senses. A discussion on complexity is also provided. Numerical simulations include a synthesized-data experiment and real-data experiments on temperature prediction, brain activity estimation, and image reconstruction. Results show that the RFF-based batch implementation offers competitive performance with a reduced computational burden when compared to the conventional KRG. The MGKRG offers a convenient trade-off between performance and complexity by varying the number of mini-batch samples. The RLSKRG has a faster convergence than the MGKRG and matches the performance of the batch implementation. [less ▲]

Detailed reference viewed: 27 (3 UL)
Full Text
Peer Reviewed
See detailNB-IoT Random Access for Non-Terrestrial Networks: Preamble Detection and Uplink Synchronization
Chougrani, Houcine UL; Kisseleff, Steven UL; Alves Martins, Wallace UL et al

in IEEE Internet of Things Journal (2021)

The satellite component is recognized as a promising solution to complement and extend the coverage of future Internet of things (IoT) terrestrial networks (TNs). In this context, a study item to ... [more ▼]

The satellite component is recognized as a promising solution to complement and extend the coverage of future Internet of things (IoT) terrestrial networks (TNs). In this context, a study item to integrate satellites into narrowband-IoT (NBIoT) systems has been approved within the 3rd Generation Partnership Project (3GPP) standardization body. However, as NBIoT systems were initially conceived for TNs, their basic design principles and operation might require some key modifications when incorporating the satellite component. These changes in NB-IoT systems, therefore, need to be carefully implemented in order to guarantee a seamless integration of both TN and non-terrestrial network (NTN) for a global coverage. This paper addresses this adaptation for the random access (RA) step in NBIoT systems, which is in fact the most challenging aspect in the NTN context, for it deals with multi-user time-frequency synchronization and timing advance for data scheduling. In particular, we propose an RA technique which is robust to typical satellite channel impairments, including long delays, significant Doppler effects, and wide beams, without requiring any modification to the current NB-IoT RA waveform. Performance evaluations demonstrate the proposal’s capability of addressing different NTN configurations recently defined by 3GPP for the 5G new radio system. [less ▲]

Detailed reference viewed: 99 (12 UL)
Full Text
Peer Reviewed
See detailEnablers for Matching Demand in GEO Multi-Beam Satellites: Dynamic Beamforming, Precoding, or Both?
Chaker, Haythem UL; Maturo, Nicola UL; Chatzinotas, Symeon UL et al

Scientific Conference (2021, September 30)

In trending satellite communication applications, the traffic demand is not only rapidly increasing, it is also spatiotemporally evolving. This motivates the deployment of high throughput satellite ... [more ▼]

In trending satellite communication applications, the traffic demand is not only rapidly increasing, it is also spatiotemporally evolving. This motivates the deployment of high throughput satellite systems with flexible radio resource management and transmission techniques. In contrast to regular beam layout plans (RBLP) currently used in GEO payloads, future flexible payloads are capable of dynamic beamforming (DBF) in order to illuminate the coverage area using highly-directive and traffic-adaptive beampatterns. The beampatterns in an adaptive beam layout plan (ABLP) can have irregular shapes and mutual overlaps, potentially causing excessive inter-beam interferences (IBI) compared to the RBLP case. In this work, we evaluate the combination of DBF and precoding as the latter promises high throughputs in interference-limited conditions and is supported by the recent DVB-S2X norm. Under realistic non-uniform traffic patterns, we compare a typical RBLP against an ABLP in terms of their traffic matching performances with and without precoding. Through the comparisons, we show that DBF enables to significantly reduce the capacity mismatches using an ABLP that uniformly balances the demand distribution across beams. Noting that the ABLP is IBI agnostic, an unpredictable interference environment is built. In such conditions, precoding enables to reliably provide high throughputs through full frequency reuse. [less ▲]

Detailed reference viewed: 139 (54 UL)
Full Text
Peer Reviewed
See detailMulti-Antenna Data-Driven Eavesdropping Attacks and Symbol-Level Precoding Countermeasures
Mayouche, Abderrahmane UL; Alves Martins, Wallace UL; Tsinos, Christos UL et al

Poster (2021, June 21)

In this work, we consider secure communications in wireless multi-user (MU) multiple-input single-output (MISO) systems with channel coding in the presence of a multi-antenna eavesdropper (Eve), who is a ... [more ▼]

In this work, we consider secure communications in wireless multi-user (MU) multiple-input single-output (MISO) systems with channel coding in the presence of a multi-antenna eavesdropper (Eve), who is a legit user trying to eavesdrop other users. In this setting, we exploit machine learning (ML) tools to design soft and hard decoding schemes by using precoded pilot symbols as training data. The proposed ML frameworks allow an Eve to determine the transmitted message with high accuracy. We thereby show that MU-MISO systems are vulnerable to such eavesdropping attacks even when relatively secure transmission techniques are employed, such as symbol-level precoding (SLP). To counteract this attack, we propose two novel SLP-based schemes that increase the bit-error rate at Eve by impeding the learning process. We design these two security-enhanced schemes to meet different requirements regarding runtime, security, and power consumption. Simulation results validate both the ML-based eavesdropping attacks as well as the countermeasures, and show that the gain in security is achieved without affecting the decoding performance at the intended users. [less ▲]

Detailed reference viewed: 64 (8 UL)
Full Text
Peer Reviewed
See detailA fault detector/classifier for closed-ring power generators using machine learning
Quintanilha, Igor M.; Elias, Vitor R. M.; Silva, Felipe B. et al

in Reliability Engineering and System Safety (2021)

Condition-based monitoring of power-generation systems is naturally becoming a standard approach in industry due to its inherent capability of fast fault detection, thus improving system efficiency and ... [more ▼]

Condition-based monitoring of power-generation systems is naturally becoming a standard approach in industry due to its inherent capability of fast fault detection, thus improving system efficiency and reducing operational costs. Most such systems employ expertise-reliant rule-based methods. This work proposes a different framework, in which machine-learning algorithms are used for detecting and classifying several fault types in a power-generation system of dynamically positioned vessels. First, principal component analysis is used to extract relevant information from labeled data. A random-forest algorithm then learns hidden patterns from faulty behavior in order to infer fault detection from unlabeled data. Results on fault detection and classification for the proposed approach show significant improvement on accuracy and speed when compared to results from rule-based methods over a comprehensive database. [less ▲]

Detailed reference viewed: 39 (0 UL)
Full Text
Peer Reviewed
See detailRadio Frequency Interference Detection Using Nonnegative Matrix Factorization
Silva, Felipe B.; Cetin, Ediz; Alves Martins, Wallace UL

in IEEE Transactions on Aerospace and Electronic Systems (2021)

This work proposes a new pre-correlation interference detection technique based on nonnegative matrix factorization (NMF) for global navigation satellite system (GNSS) signals. The proposed technique uses ... [more ▼]

This work proposes a new pre-correlation interference detection technique based on nonnegative matrix factorization (NMF) for global navigation satellite system (GNSS) signals. The proposed technique uses NMF to extract the time and frequency properties of the received signal from its spectrogram. The estimated spectral shape is then compared with the spectrogram’s time slices by means of a similarity function to detect the presence of radio frequency interference (RFI). In the presence of RFI, the NMF estimated spectral shape tends to be well-defined, resulting in high similarity levels. In contrast, in the absence of RFI, the received signal is solely comprised of noise and GNSS signals resulting in a noise like spectral shape estimate, leading to considerably reduced similarity levels. The proposal exploits this different similarity levels to detect the presence of interference. Simulation results indicate that the proposed technique yields increased detection capability with low false alarm rate even in low jammer-to-noise ratio environments for both narrow and wideband interference sources without requiring fine-tuning of parameters for specific RFI types. In addition, the proposal has reduced computational complexity, when compared with an existing statistical-based detector. [less ▲]

Detailed reference viewed: 63 (1 UL)
Full Text
Peer Reviewed
See detailData-driven Precoded MIMO Detection Robust to Channel Estimation Errors
Mayouche, Abderrahmane UL; Alves Martins, Wallace UL; Chatzinotas, Symeon UL et al

in IEEE Open Journal of the Communications Society (2021)

We study the problem of symbol detection in downlink coded multiple-input multiple-output (MIMO) systems with precoding and without the explicit knowledge of the channel-state information (CSI) at the ... [more ▼]

We study the problem of symbol detection in downlink coded multiple-input multiple-output (MIMO) systems with precoding and without the explicit knowledge of the channel-state information (CSI) at the receiver. In this context, we investigate the impact of imperfect CSI at the transmitter (CSIT) on the detection performance. We first model the CSIT degradation based on channel estimation errors to investigate its impact on the detection performance at the receiver. To mitigate the effect of CSIT deterioration at the latter, we propose learning based techniques for hard and soft detection that use downlink precoded pilot symbols as training data. We note that these pilots are originally intended for signal-to-interference-plus-noise ratio (SINR) estimation. We validate the approach by proposing a lightweight implementation that is suitable for online training using several state-of-the-art classifiers. We compare the bit error rate (BER) and the runtime complexity of the proposed approaches where we achieve superior detection performance in harsh channel conditions while maintaining low computational requirements. Specifically, numerical results show that severe CSIT degradation impedes the correct detection when a conventional detector is used. However, the proposed learning-based detectors can achieve good detection performance even under severe CSIT deterioration, and can yield 4-8 dB power gain for BER values lower than 10-4 when compared to the classic linear minimum mean square error (MMSE) detector. [less ▲]

Detailed reference viewed: 82 (11 UL)
Full Text
Peer Reviewed
See detailA Novel Learning-based Hard Decoding Scheme and Symbol-Level Precoding Countermeasures
Mayouche, Abderrahmane UL; Alves Martins, Wallace UL; Tsinos, Christos G. et al

in IEEE Wireless Communications and Networking Conference (WCNC), Najing 29 March to 01 April 2021 (2021)

In this work, we consider an eavesdropping scenario in wireless multi-user (MU) multiple-input single-output (MISO) systems with channel coding in the presence of a multi-antenna eavesdropper (Eve). In ... [more ▼]

In this work, we consider an eavesdropping scenario in wireless multi-user (MU) multiple-input single-output (MISO) systems with channel coding in the presence of a multi-antenna eavesdropper (Eve). In this setting, we exploit machine learning (ML) tools to design a hard decoding scheme by using precoded pilot symbols as training data. Within this, we propose an ML framework for a multi-antenna hard decoder that allows an Eve to decode the transmitted message with decent accuracy. We show that MU-MISO systems are vulnerable to such an attack when conventional block-level precoding is used. To counteract this attack, we propose a novel symbol-level precoding scheme that increases the bit-error rate at Eve by obstructing the learning process. Simulation results validate both the ML-based attack as well as the countermeasure, and show that the gain in security is achieved without affecting the performance at the intended users. [less ▲]

Detailed reference viewed: 111 (3 UL)
Full Text
Peer Reviewed
See detailUser Selection based on Inter-channel Interference for Massive MIMO under Line-of-sight Propagation
Chaves, Rafael S.; Cetin, Ediz; Lima, Markus V. S. et al

in URSI GASS 2021, Rome 28 August - 4 September 2021 (2021)

Massive multiple-input multiple-output (MIMO) is seen as a key enabler for next-generation wireless communication systems. Increased throughput afforded by massive MIMO, however, may severely degrade when ... [more ▼]

Massive multiple-input multiple-output (MIMO) is seen as a key enabler for next-generation wireless communication systems. Increased throughput afforded by massive MIMO, however, may severely degrade when the number of users served by a single base station increases, calling for user scheduling algorithms. To deal with this problem, a new user selection algorithm, called inter-channel interference-based selection (ICIBS), is proposed. ICIBS drops those users that induce the highest interference to the remaining users. Simulations show that selecting users with ICIBS significantly improves the throughput, outperforming state-of-the-art user selection algorithms. [less ▲]

Detailed reference viewed: 63 (8 UL)
Full Text
Peer Reviewed
See detailSymbol-Level Precoding with Constellation Rotation in the Finite Block Length Regime
Kisseleff, Steven UL; Alves Martins, Wallace UL; Chatzinotas, Symeon UL et al

in IEEE Communications Letters (2021)

This paper tackles the problem of optimizing the parameters of a symbol-level precoder for downlink multiantenna multi-user systems in the finite block length regime. Symbol-level precoding (SLP) is a non ... [more ▼]

This paper tackles the problem of optimizing the parameters of a symbol-level precoder for downlink multiantenna multi-user systems in the finite block length regime. Symbol-level precoding (SLP) is a non-linear technique for multiuser wireless networks, which exploits constructive interference among co-channel links. Current SLP designs, however, implicitly assume asymptotically infinite blocks, since they do not take into account that the design rules for finite and especially short blocks might significantly differ. This paper fills this gap by introducing a novel SLP design based on discrete constellation rotations. The rotations are the added degree of freedom that can be optimized for every block to be transmitted, for instance, to save transmit power. Numerical evaluations of the proposed method indicate substantial power savings, which might be over 99% compared to the traditional SLP, at the expense of a single additional pilot symbol per block for constellation de-rotation. [less ▲]

Detailed reference viewed: 100 (8 UL)
Full Text
Peer Reviewed
See detailDME Interference Mitigation for GNSS Receivers via Nonnegative Matrix Factorization
Silva, Felipe B.; Cetin, Ediz; Alves Martins, Wallace UL

in URSI GASS 2021, Rome 28 August - 4 September 2021 (2021)

In this work a nonnegative matrix factorization based approach is proposed to mitigate the impact of interference due to distance measurement equipment (DME) signals in global navigation satellite system ... [more ▼]

In this work a nonnegative matrix factorization based approach is proposed to mitigate the impact of interference due to distance measurement equipment (DME) signals in global navigation satellite system (GNSS) receivers. The proposed approach operates by separating the DME and GNSS signals, and the results show that it outperforms the traditional pulse-blanking based techniques in terms of acquisition and carrier-to-noise ratio metrics without discarding any of the received signal samples. [less ▲]

Detailed reference viewed: 55 (1 UL)
Full Text
Peer Reviewed
See detailRobust Passive Coherent Location via Nonlinearly Constrained Least Squares
Nicolalde-Rodríguez, Daniel P.; Apolinário Jr., José A.; Alves Martins, Wallace UL

in 12th IEEE Latin America Symposium on Circuits and System (LASCAS), Arequipa 21-24 February 2021 (2021)

This paper addresses the problem of target location by means of a passive radar. Existing approaches based on time difference-of-arrival (TDOA) measurements, namely spherical interpolation and spherical ... [more ▼]

This paper addresses the problem of target location by means of a passive radar. Existing approaches based on time difference-of-arrival (TDOA) measurements, namely spherical interpolation and spherical intersection, are revisited for the case of single transmitter and multiple receivers. The mathematical formulations of these state-of-the-art approaches do not take into account possible TDOA estimation errors, which degrade the target location performance. We extend those formulations by incorporating a nonlinear constraint into the underlying least squares problem, thus conferring robustness to the location technique against TDOA estimation errors, as corroborated by extensive numerical experiments. [less ▲]

Detailed reference viewed: 60 (1 UL)
Full Text
Peer Reviewed
See detailADS-B Signal Detection via Time-Frequency Analysis for Radio Astronomy Applications
Silva, Felipe B.; Cetin, Ediz; Alves Martins, Wallace UL

in IEEE International Symposium on Circuits and Systems (ISCAS), Daegu 22-28 May 2021 (2021)

This paper proposes a time-frequency (TF) domain technique for detecting the presence of automatic dependent surveillance-broadcast (ADS-B) interference signals in radio astronomy applications. The ... [more ▼]

This paper proposes a time-frequency (TF) domain technique for detecting the presence of automatic dependent surveillance-broadcast (ADS-B) interference signals in radio astronomy applications. The proposed technique uses a priori knowledge about the ADS-B signal’s frequency information and compares it with the received signal’s spectrogram time slices via the cosine similarity function. In the presence of ADS-B signals, the similarity levels are higher, whereas in their absence the levels are lower. Hence, the proposed approach exploits this to detect the presence of such signals. Simulation results using signals from the Parkes radio telescope show the efficacy of the proposed method in detecting the presence of ADS-B signals when compared with other classic detectors. [less ▲]

Detailed reference viewed: 39 (1 UL)
Full Text
Peer Reviewed
See detailKernel Regression on Graphs in Random Fourier Features Space
Elias, Vitor R. M.; Gogenini, Vinay C.; Alves Martins, Wallace UL et al

in IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP-2021), Toronto 6-11 June 2021 (2021)

This work proposes an efficient batch-based implementation for kernel regression on graphs (KRG) using random Fourier features (RFF) and a low-complexity online implementation. Kernel regression has ... [more ▼]

This work proposes an efficient batch-based implementation for kernel regression on graphs (KRG) using random Fourier features (RFF) and a low-complexity online implementation. Kernel regression has proven to be an efficient learning tool in the graph signal processing framework. However, it suffers from poor scalability inherent to kernel methods. We employ RFF to overcome this issue and derive a batch-based KRG whose model size is independent of the training sample size. We then combine it with a stochastic gradient-descent approach to propose an online algorithm for KRG, namely the stochastic-gradient KRG (SGKRG). We also derive sufficient conditions for convergence in the mean sense of the online algorithms. We validate the performance of the proposed algorithms through numerical experiments using both synthesized and real data. Results show that the proposed batch-based implementation can match the performance of conventional KRG while having reduced complexity. Moreover, the online implementations effectively learn the target model and achieve competitive performance compared to the batch implementations. [less ▲]

Detailed reference viewed: 73 (4 UL)
Full Text
Peer Reviewed
See detailDiffusion-based Virtual Graph Adjacency for Fourier Analysis of Network Signals
Elias, Vitor R. M.; Alves Martins, Wallace UL; Werner, Stefan

in XXXVIII SIMPÓSIO BRASILEIRO DE TELECOMUNICAÇÕES E PROCESSAMENTO DE SINAIS, Florianópolis 22-25 November 2020 (2020, November)

This work proposes a graph model for networks where node collaborations can be described by the Markov property. The proposed model augments an initial graph adjacency using diffusion distances. The ... [more ▼]

This work proposes a graph model for networks where node collaborations can be described by the Markov property. The proposed model augments an initial graph adjacency using diffusion distances. The resulting virtual adjacency depends on a diffusion-scale parameter, which leads to a controlled shift in the graph-Fourier-transform spectrum. This enables a frequency analysis tailored to the actual network collaboration, revealing more information on the graph signal when compared to traditional approaches. The proposed model is employed for anomaly detection in real and synthetic networks, and results confirm that using the proposed virtual adjacency yields better classification than the initial adjacency. [less ▲]

Detailed reference viewed: 119 (5 UL)
Full Text
Peer Reviewed
See detailOversampled DFT-Modulated Biorthogonal Filter Banks: Perfect Reconstruction Designs and Multiplierless Approximations
Alves Martins, Wallace UL; Shankar, Bhavani UL; Ottersten, Björn UL

in IEEE Transactions on Circuits and Systems. II, Express Briefs (2020), 67(11), 2777-2781

We propose a novel methodology for designing oversampled discrete Fourier transform-modulated uniform filter banks. The analysis prototype is designed as a Nyquist filter, whereas the synthesis prototype ... [more ▼]

We propose a novel methodology for designing oversampled discrete Fourier transform-modulated uniform filter banks. The analysis prototype is designed as a Nyquist filter, whereas the synthesis prototype is designed to guarantee perfect reconstruction (PR) considering oversampling. The resulting optimization problem fits into the disciplined convex programming framework, as long as some convex objective function is employed, as the minimization of either the stop-band energy or the maximum deviation from a desired response. The methodology also accounts for near-PR multiplierless approximations of the prototype analysis and synthesis filters, whose coefficients are obtained in the sum-of-power-of-two (SOPOT) space. The quantized coefficients are computed using successive approximation of vectors, which is able to yield filters with a reduced number of SOPOT coefficients in a computationally efficient manner. The proposed methodology is especially appealing for supporting actual hardware deployments, such as modern digital transparent processors to be used in next-generation satellite payloads. [less ▲]

Detailed reference viewed: 265 (27 UL)
Full Text
Peer Reviewed
See detailGraph Diffusion Kernel LMS using Random Fourier Features
Gogineni, Vinay; Elias, Vitor R. M.; Alves Martins, Wallace UL et al

in 2020 54th Asilomar Conference on Signals, Systems, and Computers, Pacific Grove, CA, USA, 1-5 November 2020 (2020, November)

This work introduces kernel adaptive graph filters that operate in the reproducing kernel Hilbert space. We propose a centralized graph kernel least mean squares (GKLMS) approach for identifying the ... [more ▼]

This work introduces kernel adaptive graph filters that operate in the reproducing kernel Hilbert space. We propose a centralized graph kernel least mean squares (GKLMS) approach for identifying the nonlinear graph filters. The principles of coherence-check and random Fourier features (RFF) are used to reduce the dictionary size. Additionally, we leverage on the graph structure to derive the graph diffusion KLMS (GDKLMS). The proposed GDKLMS requires only single-hop communication during successive time instants, making it viable for real-time network-based applications. In the distributed implementation, usage of RFF avoids the requirement of a centralized pretrained dictionary in the case of coherence-check. Finally, the performance of the proposed algorithms is demonstrated in modeling a nonlinear graph filter via numerical examples. The results show that centralized and distributed implementations effectively model the nonlinear graph filters, whereas the random feature-based solutions is shown to outperform coherence-check based solutions. [less ▲]

Detailed reference viewed: 121 (6 UL)