Call for Papers
The workshop will primarily focus on addressing the challenges encountered in deploying and developing edge computing applications using the serverless paradigm. As we know, serverless computing, also known as Function as a Service (FaaS), has gained significant momentum in recent times and is deemed the next-generation cloud service delivery paradigm. Serverless computing will help users, and developers smoothly and seamlessly deploy their applications for Edge computing environments like sensor-based technologies, the Internet of Things, cyber-physical systems, big data analytics, machine learning, cognitive computing, and artificial intelligence. All the service providers of the major league, i.e., Amazon, Google, Microsoft, have successfully launched commercially usable serverless computing platforms. Although the services are commercially available and functional, several unresolved and challenging issues persist.
This workshop aims to provide a forum for researchers and practitioners to exchange innovative ideas, latest research findings, practical experiences, lessons learned, and future directions to propel the research on utilizing serverless infrastructure for smart and pervasive computing. The topics include, but are not limited to:
- Edge-Serverless for middleware applications
- Opportunities in Serverless Edge Computing
- Stateless Life-cycle
- Cloud to Edge Integration
- Architecture support for edge computing
- Fine-grained Auto-scaling
- Faster response sensitive applications
- Edge computing in IoT offloading
- Workload characterization and analysis at the edge
- On-device artificial intelligence
- Audio/video streaming techniques
- Real-time multimedia techniques
- Quality of Service (QoS) improvements techniques
- Edge-Serverless coordinated issues
- Cold Starts handling
- Energy efficient edge computing
- Sourav Kanti Addya, National Institute of Technology Karnataka, India.
- Eirini Eleni Tsiropoulou, University of New Mexico Albuquerque, NM, USA.
- Sandip Chakraborty, Indian Institute of Technology Kharagpur, India.
Call for Papers
Intent-Based Networking (IBN) is a new technology that aims to automate the configuration of networks and facilitate the interaction of all-levels of users with the underlying infrastructure and network equipment. IBN, through the submission of high-level intents describing what the user wants instead of how to do it, is expected to create a flexible and vendor agnostic manner to manage the network, while reducing the complexity and the manually error-prone process to configure the network.
IBN is creating immense opportunities for any kind of network operator and infrastructure provider and reduces the knowledge gap between novice and expert users. However, IBN as a relatively new technology presents a multitude of open challenges that need to be addressed. For instance, how can an intent be expressed appropriately by network users of different experience and addressed to networks with different capabilities? Furthermore, how can intents be accurately translated into low-level configuration policies? Additionally, how can the intent activation be conducted efficiently and be assured throughout its lifetime to guarantee a high- performance configuration that will present the necessary self-properties of an autonomous network?
The above are just few of the many challenges that IBN has to tackle. Accordingly, this workshop is soliciting conceptual, theoretical and experimental contributions to a set of currently unresolved challenges in the area of IBN and autonomous networks, with the goal to create High Performance and self-managed networks.
- Intent expression through Natural Language Processing
- Intent modeling and policy extraction
- IBN specific languages
- Automated policy conflict detection and resolution
- Intent activation towards high performance networks
- Intent assurance for self-managed networks (Self-Healing, Self-Configuration, Self-
Optimization, Self-Protection, etc.)
- Architectural considerations for management and orchestration of IBN Systems
- Performance analysis of IBN Systems
- Business and techno-economics opportunities for IBN applications and use cases
- Marios Avgeris, Carleton University, Ottawa, Canada.
- Aris Leivadeas, École de Technologie Supérieure (ÉTS), Montreal, Canada.
Call for Papers
Machine learning (ML) systems are gaining immense popularity and are increasingly deployed in edge computing devices and networks. These devices are characterized by limited computing capabilities, being also restricted by power constraints. Similarly, edge-first networks face delays, jitter, and packet losses due to resource contention, high traffic loads, and other reasons. ML can be used to process and analyze data from edge devices and sensors to extract useful information and insights. This information can then be used to make decisions about how to manage and optimize at run-time both the edge devices and the network. Additionally, ML can help to identify patterns and correlations in data, which can be used to improve decision-making. However, there are a number of challenges associated with implementing machine learning on edge devices and networks.
The RCML Workshop aims to stimulate research on the latest advancements in resource-constraint machine learning for edge systems. Research results from funded projects in the general area of machine learning optimizations for edge computing are especially encouraged. Overall, the workshop seeks original manuscripts in the scope of the workshop, but not limited to:
- Resource-aware machine learning algorithms
- Energy-efficient hardware accelerators for machine learning
- Machine learning for edge-first networks
- Resource management in edge-first networks
- Approximate computing
- Reconfigurable systems
- Methods for machine learning optimization and compression
- Emerging design technologies for future computing
- Power-efficient and sustainable computing
- Internet of Things
- Case studies of machine learning for edge systems
- Protocols for communication in edge-first networks
- End-to-end protocols, flow and congestion control
- Pervasive and wearable computing and networking
- Artificial intelligence and machine learning for wireless networks
- Attack modelling, prevention, mitigation, and defense in wireless networks
- Reinforcement learning and deep learning for networks
- Iraklis Anagnostopoulos, Southern Illinois University Carbondale, USA.
- Paper Submission Due: April 15, 2023
- Acceptance Notification: April 21, 2023
- Author Registration Deadline: May 5, 2023
- Final Version Submission Due: May 5, 2023