Video content is responsible for more than 70% of the global IP traffic. Consequently, it is important for content delivery infrastructures to rapidly detect and respond to changes in content popularity dynamics. In this paper, we propose the employment of on-line change point (CP) analysis to implement real-time, autonomous and low-complexity video content popularity detection. Our proposal, denoted as real-time change point detector (RCPD), estimates the existence, the number and the direction of changes on the average number of video visits by combining: (i) off-line and on-line CP detection algorithms; (ii) an improved time-series segmentation heuristic for the reliable detection of multiple CPs; and (iii) two algorithms for the identification of the direction of changes. The proposed detector is validated against synthetic data, as well as a large database of real YouTube video visits. It is demonstrated that the RCPD can accurately identify changes in the average content popularity and the direction of change. In particular, the success rate of the RCPD over synthetic data is shown to exceed 94% for medium and large changes in content popularity. Additionally, the dynamic time warping distance, between the actual and the estimated changes, has been found to range between 20 samples on average, over synthetic data, to 52 samples, in real data. The rapid responsiveness of the RCPD is instrumental in the deployment of real-time, lightweight load balancing solutions, as shown in a real example.

As video content is responsible for more than 70% of the global IP traffic, related resource allocation approaches, e.g., using content caching, become increasingly important. In this context, to avoid under-provisioning, it is important to rapidly detect and respond to changes in content popularity dynamics, including volatility, i.e., changes in the second order moment of the underlying process. In this paper, we focus on the early identification of changes in the variance of video content popularity, which we address as a statistical change point (CP) detection problem. Unlike changes in the mean that can be well captured by non-parametric statistical approaches, to address this more demanding problem, we construct a hypothesis test that uses in the test statistic both parametric and non-parametric approaches. In the context of parametric models, we consider linear, in the form of autoregressive moving average (ARMA), and, nonlinear, in the form of generalized autoregressive conditional heteroskedasticity (GARCH) processes. We propose an integrated algorithm that combines off-line and on-line CP schemes, with the off-line scheme used as a training (learning) phase. The algorithm is first assessed over synthetic data; our analysis demonstrates that non parametric and GARCH model based approaches can better generalize and are better suited for content views time series with unknown statistics. Finally, the non-parametric and the GARCH based variations of our proposed integrated algorithm are applied on real YouTube video content views time series, to illustrate the performance of the proposed approach of volatility change detection.

Software-defined wireless mesh networks are being increasingly deployed in diverse settings, such as smart cities and public Wi-Fi access infrastructures. The signal propagation and interference issues that typically characterize these environments can be handled by employing SDN controller mechanisms, effectively monitoring link quality and triggering appropriate mitigation strategies, such as adjusting link and/or routing protocols. In this paper, we propose an unsupervised machine learning (ML) online framework for link quality detection consisting of: 1) improved preprocessing clustering algorithm, based on elastic similarity measures, to efficiently characterize wireless links in terms of reliability, and 2) a novel change point (CP) detector for the real-time identification of anomalies in the quality of selected links, which minimizes the overestimation error through the incorporation of a rank-based test and a recursive max-type procedure. In this sense, considering the communication constraints of such environments, our approach minimizes the detection overhead and the inaccurate decisions caused by overestimation. The proposed detector is validated, both on its individual components and as an overall mechanism, against synthetic but also real data traces; the latter being extracted from real wireless mesh network deployments.

Smart cities are progressively emerging and require city-wide networks able to maintain low delay and increased reliability. However, wireless communications in the urban environment are occasionally disrupted, leading to volatile network conditions and impeding traditional networks to cope with these demands. To this end, named data networking (NDN) and delay/disruption tolerant networking (DTN) architectures, as well as the combined NDN-over-DTN (NoD) scheme have been introduced and are actively investigated for combating distinct communication challenges in wireless city-wide deployments. In this work we investigate the performance and reliability tradeoffs of each approach in the smart city context, and accordingly propose an experimental system that employs interchangeably NDN, DTN and NoD driven by a novel change point (CP)-based network reliability detector. The real-world experiments showed that the proposed multiprotocol adaptive system, called ALTER, leads to lower latency and increased reliability of the urban network, compared to nonadaptive NDN and NoD solutions.

The Internet of Things (IoT) is a key technology for smart community networks, such as smart-city environments, and its evolution calls for stringent performance requirements (e.g., low delay) to support efficient communication among a wide range of objects, including people, sensors, vehicles, etc. At the same time, these ecosystems usually adopt wireless mesh technology to extend their communication range in large-scale IoT deployments. However, due to the high range of coverage, the smart-city WMNs may face different network challenges according to the network characteristic, for example, (i) areas that include a significant number of wireless nodes or (ii) areas with frequent dynamic changes such as link failures due to unstable topologies. Named-Data Networking (NDN) can enhance WMNs to meet such IoT requirements, thanks to the content naming scheme and in-network caching, but it necessitates adaptability to the challenging conditions of WMNs. In this work, we aim at efficient end-to-end NDN communication in terms of performance (i.e., delay), performing extended experimentation over a real WMN, evaluating and discussing the benefits provided by two SDN-based NDN strategies: (1) a dynamic SDN-based solution that integrates the NDN operation with the routing decisions of a WMN routing protocol; (2) a static one which based on SDN-based clustering and real WMN performance measurements. Our key contributions include (i) the implementation of two types of NDN path selection strategies; (ii) experimentation and data collection over the w-iLab.t Fed4FIRE+ testbed with real WMN conditions; (ii) real measurements released as open-data, related to the performance of the wireless links in terms of RSSI, delay, and packet loss among the wireless nodes of the corresponding testbed.

While the literature on channel state information (CSI)-based authentication and key distillation is vast, the two topics have customarily been studied separately. This paper proposes unsupervised learning techniques to disentangle deterministic from stochastic fading to decompose observed CSI vectors into “predictable” and “unpredictable” components. The former, primarily due to large-scale fading, can be used for node authentication. The latter, primarily due to small-scale fading, can be used for secret key generation (SKG). The parameterization of the decomposition is performed using the following metrics: (i) a CSI fingerprint “separability” criterion, expressed through the maximisation of the total variation distance (TVD) between the empirical CSI fingerprints; (ii) a statistical independence metric for CSI collected at different users in neighboring locations, using the d-dimensional Hilbert Schmidt independence criterion (dHSIC) test statistic; (iii) an estimation of information leakage at different users to determine the amount of necessary hashing for privacy amplification in the SKG using the FBLAEU machine learning based conditional min-entropy estimator. Employing principal component analysis (PCA), kernel PCA and autoencoders on synthetic and natural CSI datasets, this work shows that explicit security guarantees can be provided by using physical layer security for authentication and key agreement.