Information Theory (cs.IT)

Abstract: The capabilities of machine intelligence are bounded by the potential of data from the past to forecast the future. Deep learning tools are used to find structures in the available data to make predictions about the future. Such structures have to be present in the available data in the first place and they have to be applicable in the future. Forecast ergodicity is a measure of the ability to forecast future events from data in the past. We model this bound by the algorithmic complexity of the available data.

Abstract: To obviate the control of reflective intelligent surfaces (RISs) and the related control overhead, recent works envisioned autonomous and self-configuring RISs that do not need explicit use of control channels. Instead, these devices, named hybrid RISs (HRISs), are equipped with receiving radio-frequency (RF) chains and can perform sensing operations to act independently and in parallel to the other network entities. A natural problem then emerges: as the HRIS operates concurrently with the communication protocols, how should its operation modes be scheduled in time such that it helps the network while minimizing any undesirable effects? In this paper, we propose an orchestration framework that answers this question revealing an engineering trade-off, called the self-configuring trade-off, that characterizes the applicability of self-configuring HRISs under the consideration of massive multiple-input multiple-output (mMIMO) networks. We evaluate our proposed framework considering two different HRIS hardware architectures, the power- and signal-based HRISs that differ in their hardware complexity. The numerical results show that the self-configuring HRIS can offer significant performance gains when adopting our framework.

Abstract: This paper considers an anomaly detection problem in which a detection algorithm assigns anomaly scores to multi-dimensional data points, such as cellular networks' Key Performance Indicators (KPIs). We propose an optimization framework to refine these anomaly scores by leveraging side information in the form of a causality graph between the various features of the data points. The refinement block builds on causality theory and a proposed notion of confidence scores. After motivating our framework, smoothness properties are proved for the ensuing mathematical expressions. Next, equipped with these results, a gradient descent algorithm is proposed, and a proof of its convergence to a stationary point is provided. Our results hold (i) for any causal anomaly detection algorithm and (ii) for any side information in the form of a directed acyclic graph. Numerical results are provided to illustrate the advantage of our proposed framework in dealing with False Positives (FPs) and False Negatives (FNs). Additionally, the effect of the graph's structure on the expected performance advantage and the various trade-offs that take place are analyzed.