Beyond the Mobile-Cloud Computing Paradigm

Computing is changing from a pure elastic provisioning of virtual resources (or platforms) to a transparent and adaptive hosting environment that fully realizes the “everything as a service” provisioning concept, from centralised cloud to the edge, and from network and computing infrastructure up to the application layers. While cloud-based solutions have gained wide adoption, key drivers for the recent move towards the edge can be listed as follows: (i) a constant and unprecedented increase in the mobile data traffic puts a stress on the mobile network operators which strive for traffic reduction and/or better management of their resources, (ii) emergence of novel latency-sensitive applications such as AR/VR or real-time analytics making the traditional cloud-based solutions undesirable, (iii) along with latency and data traffic considerations, desire to keep the sensitive data local at a protected region, e.g., smart industry scenarios. Edge computing aims at mitigating these challenges by moving some parts of  service-specific computation and data storage to the proximity of the devices that generate the data. For instance, the edge can process the data so that only the most valuable or needed information is transmitted to the cloud. Despite these mentioned motivations, edge paradigm itself might exhibit various limitations given that the edge is expected to have less resources compared to the well-provisioned resource-rich cloud. Such limitations become more visible especially for some Internet-of-Things (IoT) devices that are lightweight by design. Given that these devices will become an integral part of the networked systems, e.g., smart logistics or smart cities, it is crucial to develop schemes that can exploit resources available at the edge or beyond. Similarly, many questions remain yet to be addressed regarding deployment, security, resource management, and privacy aspects. 

This workshop will explore how we distribute computing, storage, and ML executions across heterogeneous resources between cloud and end-user devices to strike a balance between the performance and user-centric factors such as data privacy and explainability. Two key research questions this workshop intends to tackle for these listed directions are as follows:  

• Q1: How can we exploit the available resources distributed in the continuum of edge to cloud to serve emerging applications, including time-critical and safety-critical ones, with high efficiency while ensuring various design goals such as privacy protection or reliability?  
• Q2: How can we include security-by-design and privacy-by-design rather than these two crucial aspects being treated as an afterthought? 

From scientific and community perspectives, the aims of our workshop include: 

• Creating a scientific community within the Netherlands of like-minded but also complementary-skilled researchers who are motivated to tackle the unique challenges of edge computing and unlocking the value of edge computing in research and education. Our workshop will serve to create a discussion platform which can be continued after the event through digital platforms such as slack channel or a mailing list and potentially leading to an annual event in the future. 
• Building strong ties of collaboration with the leading researchers in edge computing in Europe and beyond.   
• Developing a joint research agenda and identifying unique challenges for which a higher critical mass and/or interdisciplinary team is necessary in order to address them at scale (Dutch and international level). 

 To go beyond the conventional mobile-cloud paradigm, the new Edge Computing field is essentially an interdisciplinary field where we need researchers and engineers from system, networking, data/AI, security & privacy domains to develop the solutions in a coherent manner, rather than merging separate solutions together as afterthought.