Graph Signal Processing in Applications to Sensor Networks, Smart Grids, and Smart Cities

Jablonski, Ireneusz

doi:10.1109/jsen.2017.2733767

Cited by 90 publications

(63 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another area of interest is environmental monitoring. As an example, the author of [74] has proposed to apply the GSP-based graph learning framework of [70] for the analysis of exemplary environmental data of ozone concentration in Poland. More specifically, the paper has proposed to learn a network that reflects the relationship between different regions in terms of ozone concentration.…”

Section: Other Application Domainsmentioning

confidence: 99%

“…Generally speaking, inferring graph topologies from observations is an ill-posed problem, and there are many ways of associating a topology with the observed data samples. Some of the most straightforward May 21, 2019 DRAFT methods include computing sample correlation, or using a similarity function, e.g., a Gaussian RBF kernel function, to quantify the similarity between data samples. These methods are based purely on observations without any explicit prior or model of the data, hence they may be sensitive to noise and have difficulty in tuning the hyper-parameters.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Learning Graphs From Data: A Signal Representation Perspective

Dong

Thanou

Rabbat

et al. 2019

IEEE Signal Process. Mag.

353

222

View full text Add to dashboard Cite

The construction of a meaningful graph topology plays a crucial role in the effective representation, processing, analysis and visualization of structured data. When a natural choice of the graph is not readily available from the data sets, it is thus desirable to infer or learn a graph topology from the data. In this tutorial overview, we survey solutions to the problem of graph learning, including classical viewpoints from statistics and physics, and more recent approaches that adopt a graph signal processing (GSP) perspective. We further emphasize the conceptual similarities and differences between classical and GSP-based graph inference methods, and highlight the potential advantage of the latter in a number of theoretical and practical scenarios. We conclude with several open issues and challenges that are keys to the design of future signal processing and machine learning algorithms for learning graphs from data.

show abstract

Section: Other Application Domainsmentioning

confidence: 99%

mentioning

confidence: 99%

Learning Graphs From Data: A Signal Representation Perspective

Dong

Thanou

Rabbat

et al. 2019

IEEE Signal Process. Mag.

353

222

View full text Add to dashboard Cite

show abstract

“…, e N N } denotes the set of edges. Each vertex corresponds to one element of the network [3], [4], [6], [10], [19], [20], [33]. Each edge e ij represents a pairwise connection between nodes v i and v j .…”

Section: A Graphs and Network Modelingmentioning

confidence: 99%

“…L ARGE quantities of heterogeneous data are constantly collected by numerous sensors, which are often geographically dispersed. Real networks and their corresponding data come in vastly different shapes and applications, ranging from genetic interaction networks [1] and the human brain [2] to sensor networks and smart cities [3]. The increased connectivity and availability of abundant data calls for methods that can uncover hidden connections between seemingly unrelated things in complex and irregular structures.…”

Section: Introductionmentioning

confidence: 99%

Extended Adjacency and Scale-Dependent Graph Fourier Transform via Diffusion Distances

Elias

Martins

Werner

2020

IEEE Trans. on Signal and Inf. Process. over Networks

View full text Add to dashboard Cite

This paper proposes the augmentation of the adjacency model of networks for graph signal processing. It is assumed that no information about the network is available, apart from the initial adjacency matrix. In the proposed model, additional edges are created according to a Markov relation imposed between nodes. This information is incorporated into the extended-adjacency matrix as a function of the diffusion distance between nodes. The diffusion distance measures similarities between nodes at a certain diffusion scale or time, and is a metric adopted from diffusion maps. Similarly, the proposed extendedadjacency matrix depends on the diffusion scale, which enables the definition of a scale-dependent graph Fourier transform. We conduct theoretical analyses of both the extended adjacency and the corresponding graph Fourier transform and show that different diffusion scales lead to different graph-frequency perspectives. At different scales, the transform discriminates shifted ranges of signal variations across the graph, revealing more information on the graph signal when compared to traditional approaches. The scale-dependent graph Fourier transform is applied for anomaly detection and is shown to outperform the conventional graph Fourier transform.

show abstract

“…The "smart city" is a combination of modern technologies and solutions, which allows improving the quality of life of citizens and, at the same time, help the city economy more efficiently use available resources [8]. The progress of the Smart Cities is based on the introduction of new ways, when, thanks to sensors and algorithms, the control and management capabilities are enhanced [9,10]. Sensors have become indispensable in many industries because they provide vital information about parameters that include temperature, position, chemistry, pressure, force and load, flow and level, and thus affect products, processes and systems.…”

Section: Introductionmentioning

confidence: 99%

Automated system for synthesis of sensors for smart cities

et al. 2019

View full text Add to dashboard Cite

The concept of a smart city is aimed at developing technologies and infrastructure based on the use of new technologies, large data centers, smart sensors and automated power grids. As technological advances in control systems increase the efficiency of sensors, users are calling for advances in sensor accuracy, reliability, dynamic characteristics (response time, dynamic linearity etc.), robustness, miniaturization and communication capability. The sensor industry is expected to experience sound growth during the five-year forecast period. It is possible satisfy the growing demand, in particular, by using automated systems of conceptual design of sensors, which can reduce the design time several times. Physical principles of operation of the sensors involved in information measuring and control systems of buildings are quite diverse. In this article, it is proposed to use a modified energy-information method for the conceptual design of sensors, invariant to the physical principle of operation. The possibility of using the operator method for the calculation and study of the dynamic characteristics of the designed elements of control systems of different physical nature is shown. The architecture of the expert system of evaluation of dynamic characteristics of sensitive elements of information-measuring devices of smart home is developed.

show abstract

Graph Signal Processing in Applications to Sensor Networks, Smart Grids, and Smart Cities

Cited by 90 publications

References 34 publications

Learning Graphs From Data: A Signal Representation Perspective

Learning Graphs From Data: A Signal Representation Perspective

Extended Adjacency and Scale-Dependent Graph Fourier Transform via Diffusion Distances

Automated system for synthesis of sensors for smart cities

Contact Info

Product

Resources

About