Network management tools are usually inherited from one generation to another. This was successful since these tools have been kept in check and updated regularly to fit new networking goals and service requirements. Unfortunately, new networking services will render this approach obsolete and handcrafting new tools or upgrading the current ones may lead to complicated systems that will be difficult to maintain and improve. Fortunately, recent advances in AI have provided new promising tools that can help solving many network management problems. Following this interesting trend, the current article presents LEASCH, a deep reinforcement learning model able to solve the radio resource scheduling problem in the MAC layer of 5G networks. LEASCH is developed and trained in a sand-box and then deployed in a 5G network. It has been evaluated under different numerology settings. The experimental results show that it is both numerology-agnostic and efficient when compared to conventional baseline methods in many key performance indicators.

Switch-controller assignment is an essential task in multi-controller software-defined networking. Static assignments are not practical because network dynamics are complex and difficult to predetermine. Since network load varies both in space and time, the mapping of switches to controllers should be adaptive to sudden changes in the network. To that end, switch migration plays an important role in maintaining dynamic switch-controller mapping. Migrating switches from overloaded to underloaded controllers brings flexibility and adaptability to the network but, at the same time, deciding which switches should be migrated to which controllers, while maintaining a balanced load in the network, is a challenging task. This work presents a heuristic approach with solution shaking to solve the switch migration problem. Shift and swap moves are incorporated within a search scheme. Every move is evaluated by how much benefit it will give to both the immigration and outmigration controllers. The experimental results show that the proposed approach is able to outweigh the state-of-art approaches, and improve the load balancing results up to ≈ 14% in some scenarios when compared to the most recent approach. In addition, the results show that the proposed work is more robust to controller failure than the state-of-art methods.

The emergence of smarter and broader peopleoriented IoT applications and services requires interoperability at both data and knowledge levels. However, although some semantic IoT architectures have been proposed, achieving a high degree of interoperability requires dealing with a sea of non-integrated data, scattered across vertical silos. Also, these architectures do not fit into the machine-to-machine (M2M) requirements, as data annotation has no knowledge on object interactions behind arriving data. This article presents a vision of how to overcome these issues. More specifically, the semantic profiling of objects, through CoRE related standards, is envisaged as the key for data integration, allowing more powerful data annotation, validation and reasoning. These are the key blocks for the development of intelligent applications.

Wireless sensor networks (WSNs) are starting to have a high impact on our societies and, for next generation WSNs to become more integrated with the Internet, researchers recently proposed to embed IPv6 into such very constrained networks. Also, constraint application protocol (CoAP) and Observe have been proposed for RESTful services to be provided. CoAP/Observe supports the use of caches/proxies and, for this reason, an observation request may resort to multiple client/server registration steps in order to get notifications. Here, we propose to plan the multiple registration steps, at proxies, of multiple observation requests in order to make proper aggregation/scheduling of notifications for transmission. This leads to less energy consumption and to an effective use of bandwidth, avoiding energy depletion of nodes, and increasing the network lifetime. Besides, mathematically formalizing the problem, a heuristic approach is developed and a discussion on how to incorporate algorithm's decision into the network is done. The proposed framework can be applied to multiple application domains (e.g., monitoring, machine to machine).