Spain

Recent innovations and progress in technologies like wireless multimedia surveillance networks, embedded systems, cloud computing, medical diagnostics, blockchain, and Big data analytics have created numerous opportunities for IoT based applications. An enormous number of sensors deployed in IoT infrastructures like smart cities, smart homes, intelligent transportation, healthcare, and precision agriculture are a major source of highly precious big data. Lack of intelligent mechanisms for storage, indexing, retrieval, and management of sensors big data hinders efficient processing to gather actionable intelligence. Consequently, revealing concealed information and gathering actionable intelligence from IoT data is becoming increasingly sophisticated and infeasible to be accomplished with conventional Machine Learning (ML) paradigms. To this end, Deep learning (DL) can play a crucial role in the effective utilization of IoT devices, producing actionable intelligence for correct and timely decisions. Researchers around the world are motivated by the performance of DL based smart systems including computer vision, machine translation, information retrieval, fault detection, and speech recognition, etc. Due to the limited resources of IoT devices, optimization of DL models, and inferencing are investigated for resource-constrained devices in order to improve their intelligence and efficiency. Furthermore, the heterogeneous nature of underlying hardware in IoT infrastructures raise compatibility issues regarding processing and fusion of information. Hardware/software heterogeneity, data collection, data preprocessing, data generation of IoT in a massive scale, and resource-constrained nature of IoT devices are the current egoistic trends for the researchers to be investigated. In this talk, I will discuss the aforementioned aspects and will envision merging deep learning with IoT to explore new horizons for applications such as health monitoring, disease analysis, indoor localization, intelligent control, home robotics, traffic prediction, traffic monitoring, autonomous driving, and manufacture inspection, which are a few of the futuristic goals for the researchers today.

WINSYS 2021

Efficient Deep Learning Methods for IoT Applications: Current Challenges and Future Directions

The purpose of IJCCI is to bring together researchers, engineers and practitioners on the areas of Fuzzy Computation, Evolutionary Computation and Neural Computation. 

IJCCI 2021

keynote

The purpose of WINSYS, the International Conference on Wireless Networks and Mobile Systems, is to bring together researchers, engineers and practitioners interested on information systems and applications in the context of wireless networks and mobile technologies. Information systems and information technology are pervasive in the whole communications field, which is quite vast, encompassing a large number of research topics and applications: from practical issues to the more abstract theoretical aspects of communication; from low level protocols to high-level networking and applications; from wireless networking technologies to mobile information systems; many other topics are included in the scope of WINSYS.

The purpose of WINSYS, the International Conference on Wireless Networks and Mobile Systems, is to bring together researchers, engineers and practitioners interested on information systems and applications in the context of wireless networks and mobile technologies.

High-performance Networks (HPNs) have been increasingly used for big data transfer in support of large-scale scientific and business applications. The main challenge is to meet user data transfer requirements while minimizing the waste of expensive bandwidth resources. This talk presents an exploratory analysis of data transfer performance based on several years of real-life performance data in emulated and production HPNs, and the development of a machine learning-assisted performance predictor to facilitate bandwidth reservation that matches actual user needs and avoids bandwidth over-provisioning.

Fusing Intelligence into Big Data Transfer in High-performance Networks

Wireless sensor networks have been consecrated, over the years, as the integral solution to solve most of the daily problems. The field of application of the WSN ranges from basic tasks of monitoring single events to complex intelligent systems for monitoring large cities or industrial processes. IEEE 802.11 standard has been historically used to deploy this type of network. However, the intrinsic problems of this technology are relegating it to a background level in monitoring issues. In recent years, a new type of medium and long-range network has been developed for the Internet of Things: LPWAN networks (Low Power Wide Area Network) which were created to anticipate the planned obsolescence of GPRS networks. LPWAN networks are consolidating as the best solution to implement low-cost applications and to improve our daily lives. This keynote speech aims to present the main LPWAN technologies we can currently use for developing new applications that help us in our daily tasks. Special attention will be paid to the operating characteristics of LoRa/LoRaWAN that make this technology so attractive in the development of systems and applications for home and hospital environments, for monitoring in rural, urban or domestic areas, solutions for monitoring large cities or industrial processes, including very sensitive areas such as healthcare and the welfare of people. Finally, this keynote will discuss the challenges and open research issues in the area and LWPAN technology.

LPWAN Technologies: The Use Case of LoRa-based WSNs and Its Applications

6G is the next step in the evolution of mobile communication and will be a key component of the Networked Society. In particular, 5G will accelerate the development of the Virtual world. To enable connectivity for a wide range of applications and use cases, the capabilities of 6G wireless access must extend far beyond those of previous generations of mobile communications. Examples of these capabilities include ultra high data rates, ultra low latency, ultra-high reliability, energy efficiency and extreme device densities, and will be realized by the development of 5G in combination with new radio-access technologies. Therefore, this talk explains the different key technology components of 6G and from implementation to standardization.

Premium content

Efficient Deep Learning Methods for IoT Applications: Current Challenges and Future Directions

Downloads

Next from WINSYS 2021

Fusing Intelligence into Big Data Transfer in High-performance Networks

Similar lecture

Machine Learning in Non-Stationary Environments

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES