This paper proposes an automatic anomaly detection and root cause analysis model that combines Density-based Spatial Clustering of Applications with Noise (DBSCAN) and Long Short-Term Memory (LSTM) Auto-Encoder (AE). The model is trained and tested using unlabelled real-world 4G network data from a Portuguese operator. The study demonstrates the promise of DBSCAN in separating normal network traffic patterns from abnormal ones, as well as the ability of LSTM-AE to learn daily network Key Performance Indicator (KPI) behaviour and detect anomalies and their probable root causes based on reconstruction errors. Results indicate that anomalies in available network KPIs do not always result in abnormal traffic patterns, and vice versa. Consequently, it can be inferred that relying solely on traffic volumes for DBSCAN is not an ideal method to separate normal network data from abnormal data for detecting network anomalies. Additionally, the results emphasize the importance of high-quality data in terms of sampling rate and the number of KPIs, as well as the significance of data analysis in identifying patterns across different levels of mobile networks.

The Cloud Computing (CC) paradigm has risen in recent years as a solution to a need for computation and battery-constrained User Equipment (UE) to run increasingly intensive computation tasks. Nevertheless, given the centralized nature of the CC paradigm, this option introduces significant network congestion problems and unpredictable communication delays unsuitable for real-time applications. In order to cope with these problems, the Mobile Edge Computing (MEC) concept has been introduced, which proposes to bring computation resources closer to the edge of the mobile networks in a distributed way. However, given that these edge computation resources are limited, this paradigm comes with its set of challenges that need to be solved in order to make it viable. This work proposes to innovate by presenting a network management agent capable of making offloading decisions from a heterogeneous network of UEs to a heterogeneous network of MEC servers. This agent is the orchestrator of a group of 5G Small Cells (SCeNBs), enhanced with computation and storage capabilities. In order to solve this high complexity problem, an Advantage Actor-Critic (A2C) agent is implemented and tested against several baselines. The proposed solution is shown to beat the baselines by making intelligent decisions taking into account computation, battery, delay and communication constraints ignored by the baselines. The solution is also shown to be scalable, data-efficient, robust, stable and adjustable to address not only the overall system performance but also to take into account the worst-case scenario.

Directional antennas are a promising solution for improving the range of client devices and capacity of wireless networks. Unfortunately, in LP-WAN systems directional antennas tend to be both large and expensive due to operating at subGHz frequencies. However, if a client device is willing to forgo improvements in antenna gain, it is possible to realize compact and low-cost antennas that provide spatial diversity control. In this paper, we show that by increasing spatial diversity in LPWAN clients with limited (or no) client gain, we can dramatically increase overall network capacity and improve client battery life by avoiding re-transmissions. This type of directional control can also be used for hot-spot management by more effectively load balancing clients across gateways.We performed a sensitivity analysis using a combination of real hardware and simulation to explore the impact of various switchable antenna geometries on network capacity under a variety of deployment configurations. We then designed and evaluated three prototype multi-sector array clients: (1) a switchable patch antenna configuration, (2) a digital phase-shift nulling configuration, and (3) a low-cost switched PCB element phase-shift system. Each design explores a different hardware cost vs antenna beam performance operating point. We experimentally see that our real antenna beam patterns, captured in an anechoic chamber, perform in a similar manner to our simulated prediction models in terms of beam pattern and in simulation improve network capacity by up to 28% from interference isolation alone and up to 95% when offloading hot spots between four gateways. We also perform a small measurement study of how often our final design changes its configuration when deployed over multiple days on a campus testbed.

The evolution toward the so-called smart grid (SG) needs to be supported by robust and cost-efficient advanced metering infrastructure. This article focuses on technology selection and gateway placement in hybrid SGs, wherein each smart meter (SM) is equipped with either a power line communication or a long-range wide area network protocol node to connect to its serving gateway. The latter is endowed with both technologies. In particular, we introduce an analytical framework to characterize the performance–cost tradeoffs using a multicriteria optimization approach. For each technology, a set of performance objective functions (i.e., sum-rate and max–min) has been derived based on the ergodic capacity, whereas the cost is represented by SG hardware expenses. Then, the nondominated sorting genetic algorithm-II has been exploited using an appropriate chromosome structure to provide the suitable access technology for each SM along with the optimal number and the placement of the serving GWs. Extensive numerical simulations have shown that hybrid solutions provide a good interplay between performance and investment costs, spurring the need to redefine the conventional SG designs traditionally operating in single communication technologies.

This paper presents the characterization of network delays in an IEC61850 process bus substation area network, both through theoretical analysis and simulations. Several design targets were defined considering the recommendations of standards and good design practices: number of network hops, total network delay, probability of the delay being exceeded, link load, network topology and availability. An analytical delay estimation methodology is proposed, considering both the steady-state traffic and traffic resulting from a breaker failure event. A complete substation is taken as an example for characterizing the network delays, considering a star network topology. Simulations allow to obtain the cumulative distribution functions and percentile values of network delays. Results show a good agreement between the simulation and the analytical analysis. While the delay is best characterized statistically through simulation, finding the maximum network delay through simulations can be very time consuming, making the analytical analysis more suitable.

Wireless sensor and actuator networks (WSAN) constitute an emerging technology with multiple applications in many different fields. Due to the features of WSAN (dynamism, redundancy, fault tolerance, and self-organization), this technology can be used as a supporting technology for the monitoring of critical infrastructures (CIs). For decades, the monitoring of CIs has centered on supervisory control and data acquisition (SCADA) systems, where operators can monitor and control the behavior of the system. The reach of the SCADA system has been hampered by the lack of deployment flexibility of the sensors that feed it with monitoring data. The integration of a multihop WSAN with SCADA for CI monitoring constitutes a novel approach to extend the SCADA reach in a cost-effective way, eliminating this handicap. However, the integration of WSAN and SCADA presents some challenges which have to be addressed in order to comprehensively take advantage of the WSAN features. This paper presents a solution for this joint integration. The solution uses a gateway and a Web services approach together with a Web-based SCADA, which provides an integrated platform accessible from the Internet. A real scenario where this solution has been successfully applied to monitor an electrical power grid is presented.

In Wireless Multimedia Sensor Networks (WMSN), imaging and/or audio data must be delivered in time and respecting minimum reliability guarantees. While these objectives alone are usually conflicting, the resource constraints and error-prone nature of WMSN links increase the challenge even more, requiring special transport layer support such as the one offered by the Distributed Transport for Sensor Networks (DTSN) protocol. In DTSN, caching of packets at intermediate nodes contributes to significantly minimize the number of end-to-end retransmissions, but still requires packet blocks to be fully memorized at the source. In order to comply with the memory limitations of tiny nodes without compromising throughput in multi-layer data applications, a DTSN differentiated reliability extension is proposed, allowing the source to memorize only the most important packets (base layer), relying exclusively on caching to probabilistically recover the enhancement layer packets. Such extension is mainly useful to support multimedia services (e.g., imaging) and other incremental precision sensor applications. We integrated such an extension to DTSN. Extensive simulation results show that even when only a fraction of the data is stored at the source the protocol can still recover most of the enhancement data based on the caching mechanism.

Traditional transport layer protocols have been designed to perform end-to-end error control transparently to the intermediate nodes (e.g., TCP). However, the resource constraints featured by Wireless Sensor Network (WSN) require a different paradigm where intermediate nodes are able to cache packets and if possible retransmit them on-demand to avoid incurring costly end-to-end retransmissions. Lately, wireless multimedia sensor network (WMSN) is being considered as a new research area whereby WSNs are targeted for the delivery of multimedia traffic. In this paper, we propose a NACK-based repair mechanism coupled with an adaptive MAC layer retransmission scheme in order to improve the performance of the transport protocols. Specifically, our goal is to be able to reduce real-time end-to-end delay while maintaining reliability and energy efficiency in the presence of high channel error rates. Our simulation results show that the ensemble of both mechanisms provides better good-put performance while simultaneously improving energy efficiency. Furthermore, the improved protocol also achieves lower deadline miss ratios which makes it suitable for multimedia transport. While we demonstrate the effectiveness of the mechanisms by incorporating them into the basic Distributed Transport for Sensor Networks (DTSN) protocol, they are generic enough to be applicable to other WSN transport protocols.

We propose a secure distributed transport protocol for wireless sensor networks that resists against attacks on the reliability service provided by the protocol, as well as against energy depleting attacks. Our protocol is based on the Distributed Transport for Sensor Networks (DTSN) protocol, to which we add a security extension that consists in an efficient, symmetric key based authentication scheme for control packets. Besides describing the operation of our protocol, we also provide its analysis in terms of security and overhead.

This paper presents the distributed transport for sensor networks (DTSN), a novel reliable transport protocol for convergecast and unicast communications in wireless sensor networks (WSNs). In DTSN, the source completely controls the loss recovery process in order to minimize the overhead associated with control and data packets. The basic loss recovery algorithm is based on selective repeat ARQ, employing both positive and negative acknowledgements. DTSN is able to detect when all packets of a session are lost, besides scattered gaps in the packet sequence. Caching at intermediate nodes is used to avoid the inefficiency of the strictly end-to-end transport reliability TCP-like model, commonly employed in broadband networks. Reliability differentiation is achieved by means of the smart integration of partial buffering at the source, integrated with erasure coding and caching at intermediate nodes. The simulation results attest the effectiveness of both the full reliability and the reliability differentiation mechanisms in DTSN.

This paper presents a distributed algorithm for efficient Time Division Multiple Access (TDMA) slot allocation in Wireless Sensor Networks. Unlike most of the existent proposals, which try to guarantee interference-free allocation based on the n-hop criterion (which can only work with regular topologies), the presented algorithm bases its decisions on the received signal strength, allowing it to operate independently of the WSN topology. The performance of the proposed slot allocation algorithm was compared with centralized depth-first slot allocation using computer simulation. The results show that the proposed scheme is significantly faster, only paying the price of extra energy consumption during the initial network setup phase.

This paper proposes a mechanism to combine link-adaptation and transmit power control (TPC) in IEEE 802.11a WLANs. The link-adaptation and TPC algorithms try to maximize the goodput of the WLAN, while minimizing transmit power. The algorithm runs at the sender, but it relies on feedback from the receiver. The latter notifies the sender about the signal-to-interference-plus-noise-ratio experienced in previous transmissions, allowing a prediction of the channel status. An extension of the algorithm for multicast is also presented. The implementation of the proposed mechanism in IEEE 802.11a/h/e WLANs is discussed and its performance evaluated.

One of the main results of the MOICANE project has been the deployment of a distributed QoS-aware network pilot, connecting domains under different administrative control and characterized by various access technologies. This paper addresses how IP mobility and QoS were integrated for wireless access in the MOICANE architecture. The MOICANE architecture supports both the IntServ and DiffServ IP QoS models. An IP-Data Link (IP-DL) adaptation function assures independency between protocol layers as an important requirement to facilitate heterogeneity. It keeps track of active flows either based on RSVP signalling or classification of TCP/IP packet headers. IP mobility is based on a new architecture designated as terminal independent mobile IP (TIMIP), which offers transparent mobility to legacy IP terminals, thus supporting terminal heterogeneity. TIMIP includes context transfer mechanisms that can be used to support seamless mobility in either IntServ and DiffServ based wireless access networks.

This paper presents an evaluation of the QoS enhancements to the IEEE 802.11 standard, named IEEE 802.11e, currently under specification. Both the enhanced distributed coordination function (EDCF) and hybrid coordination function (HCF) modes of medium access control (MAC) operation are analysed and compared with legacy distributed coordination function (DCF) and point coordination function (PCF). Performance evaluation is attained through computer simulation of a scenario of 802.11 b/e access to an IP core network through an access point (AP) in an infrastructure WLAN.