Ricerc@Sapienza

This paper presents two dynamic and distributed clustering algorithms for Wireless Sensor Networks (WSNs). Clustering approaches are used in WSNs to improve the network lifetime and scalability by balancing the workload among the clusters. Each cluster is managed by a cluster head (CH) node. The first algorithm requires the CH nodes to be mobile: by dynamically varying the CH node positions, the algorithm is proved to converge to a specific partition of the mission area, the generalised Voronoi tessellation, in which the loads of the CH nodes are balanced.

This paper proposes a lexicographic Deep Reinforcement Learning (DeepRL)-based approach to chance-constrained Markov Decision Processes, in which the controller seeks to ensure that the probability of satisfying the constraint is above a given threshold. Standard DeepRL approaches require i) the constraints to be included as additional weighted terms in the cost function, in a multi-objective fashion, and ii) the tuning of the introduced weights during the training phase of the Deep Neural Network (DNN) according to the probability thresholds.

Efficient and truthful mechanisms to price time on remote servers/machines have been the subject of much work in recent years due to the importance of the cloud market. This paper considers online revenue maximization for a unit capacity server, when jobs are non preemptive, in the Bayesian setting: at each time step, one job arrives, with parameters drawn from an underlying distribution.