To meet the requirements of the next generation internet-of-things (NGIoT), the TERMINET platform intends to bring more intelligence at the edge of the network. To this direction, the convergence of communication and computing networks is envisioned, taking into account the trade-off among computational accuracy, delay, and efficient use of available energy. This framework is based on programmable and flexible architectures, which are mainly based on edge-centric computing, network function virtualization, and software defined networking .
A promising approach for facilitating the convergence between wireless communication and computing is mobile edge computing (MEC), according to which the connected IoT devices can execute real-time compute-intensive applications directly at the network edge. MEC is a particularly promising approach to provide the required computational performance for emerging NGIoT applications, such as the smart grids, smart industry, healthcare, and smart farming. Although the concept of MEC provides several important advantages, it creates several interesting challenges due to the fluctuation of computational and networking resources, as well as the massive number and mobility of the IoT devices. To mitigate these challenges, a novel design and orchestration framework for MEC has been developed by the team of TERMINET, which can reduce the overall delay and the energy consumption of the IoT devices . Also, the framework can properly be used to optimize the networking and computing resources when multiple objectives need to be jointly optimized.
In more detail, the corresponding research has highlighted that the efficient use of MEC depends on two interrelated factors, namely the utilized multiple access protocol and the efficient use of the computational and communication resources . More specifically, their joint optimization can lead to substantial improvement of the overall performance in terms of delay or energy consumption. To this end, a novel generalized multiple access protocol has been introduced, while all degrees-of-freedom have been taken into account for the orchestration. Also, both full and partial offloading have been considered. In full offloading, the whole task is executed at the edge server, while in partial offloading the task can be partitioned into two parts, with one executed at the device locally while the residual can be offloaded for edge execution. In both cases, novel resource allocation schemes have been developed, which are based on tools from optimization theory. The results has verified that the proposed framework is particularly promising for the NGIoT, since it can lead to substantial reduction of delay and energy consumption compared to other alternatives.
 Diamantoulakis, P., Bouzinis, P., Sarigiannidis, P., Ding, Z., & Karagiannidis, G. G. (2021). Optimal Design and Orchestration of Mobile Edge Computing with Energy Awareness. IEEE Transactions on Sustainable Computing.