The explosive development of mobile communications and networking has led to the creation of an extremely complex system,which is difficult to manage.Hence,we propose an AI-powered network framework that uses AI technologies to operate the network automatically.However,due to the separation between different mobile network operators,data barriers between diverse operators become bottlenecks to exploit the full power of AI.In this paper,we establish a mutual trust data sharing framework to break these data barriers.The framework is based on the distributed and temper-proof attributes of blockchain.We implement a prototype based on Hyperledger Fabric.The proposed system combines supervision and fine-grained data access control based on smart contracts,which provides a secure and trustless environment for data sharing.We further compare our system with existing data sharing schemes,and we find that our system provides a better functionality.

In this paper,we present a user-complaint prediction system for mobile access networks based on network monitoring data.By applying machine-learning models,the proposed system can relate user complaints to network performance indicators,alarm reports in a data-driven fashion,and predict the complaint events in a fine-grained spatial area within a specific time window.The proposed system harnesses several special designs to deal with the specialty in complaint prediction;complaint bursts are extracted using linear filtering and threshold detection to reduce the noisy fluctuation in raw complaint events.A fuzzy gridding method is also proposed to resolve the inaccuracy in verbally described complaint locations.Furthermore,we combine up-sampling with down-sampling to combat the severe skewness towards negative samples.The proposed system is evaluated using a real dataset collected from a major Chinese mobile operator,in which,events due to complaint bursts account approximately for only 0.3% of all recorded events.Results show that our system can detect 30% of complaint bursts 3 h ahead with more than 80% precision.This will achieve a corresponding proportion of quality of experience improvement if all predicted complaint events can be handled in advance through proper network maintenance.

Driven by the need of a plethora of machine learning applications,several attempts have been made at improving the performance of classifiers applied to imbalanced datasets.In this paper,we present a fast maximum entropy machine(MEM)combined with a synthetic minority over-sampling technique for handling binary classification problems with high imbalance ratios,large numbers of data samples,and medium/large numbers of features.A random Fourier feature representation of kernel functions and primal estimated sub-gradient solver for support vector machine(PEGASOS)are applied to speed up the classic MEM.Experiments have been conducted using various real datasets (including two China Mobile datasets and several other standard test datasets) with various configurations.The obtained results demonstrate that the proposed algorithm has extremely low complexity but an excellent overall classification performance (in terms of several widely used evaluation metrics)as compared to the classic MEM and some other state-of-the-art methods.The proposed algorithm is particularly valuable in big data applications owing to its significantly low computational complexity.

Mobile big data collected by mobile network operators is of interest to many research communities and industries for its remarkable values.However,such spatiotemporal information may lead to a harsh threat to subscribers’privacy.This work focuses on subscriber privacy vulnerability assessment in terms of user identifiability across two datasets with significant detail reduced mobility representation.In this paper,we propose an innovative semantic spatiotemporal representation for each subscriber based on the geographic information,termed as daily habitat region,to approximate the subscriber’s daily mobility coverage with far lesser information compared with original mobility traces.The daily habitat region is realized via convex hull extraction on the user’s daily spatiotemporal traces.As a result,user identification can be formulated to match two records with the maximum similarity score between two convex hull sets,obtained by our proposed similarity measures based on cosine distance and permutation hypothesis test.Experiments are conducted to evaluate our proposed innovative mobility representation and user identification algorithms,which also demonstrate that the subscriber’s mobile privacy is under a severe threat even with significantly reduced spatiotemporal information.

Estates,especially those of public securityrelated companies and institutes,have to protect their privacy from adversary unmanned aerial vehicles(UAVs).In this paper,we propose a reinforcement learning-based control framework to prevent unauthorized UAVs from entering a target area in a dynamic game without being aware of the UAV attack model.This UAV control scheme enables a target estate to choose the optimal control policy,such as jamming the global positioning system signals,hacking,and laser shooting,to expel nearby UAVs.A deep reinforcement learning technique,called neural episodic control,is used to accelerate the learning speed to achieve the optimal UAV control policy,especially for estates with a large area,against complicated UAV attack policies.We analyze the computational complexity for the proposed UAV control scheme and provide its performance bound,including the risk level of the estate and its utility.Our simulation results show that the proposed scheme can reduce the risk level of the target estate and improve its utility against malicious UAVs compared with the selected benchmark scheme.

WirelessHART is one of the most widely used technologies in industrial wireless networks.However,its performance is highly influenced by the quality of wireless channels.To improve the reliability of wireless communications,WirelessHART employs channel blacklisting and channel hopping mechanisms,which highlights the importance of channel assessment.Traditional methods generally resort to packet reception ratio (PRR) of the previous time slot to assess and allocate channels,but this is not accurate.In this paper,we propose a learning-based framework for predicting the PRR,and on the basis of the predicted PRR,we develop a heuristic channel selection algorithm to confirm the channel list,which takes into account the balance of channel diversity and route diversity.Simulation results demonstrate that our algorithm outperforms existing ones in terms of achieved reliability.

Ubiquitous computing facilitated by Internet of things (IoT) devices has made modern day life easier across many areas.It offers capabilities to measure parameters associated with the devices,to infer from their results,and to understand and control millions of such devices in various application domains.The enormous potential of IoT systems enables each and every device to communicate with each other,thereby providing more productivity.In this scenario,heterogeneity of technologies in use is expected to intensify the security threats.Policy enforcement for the assurance of privacy and security plays a key role in these systems.Fulfillment of privacy and security related requirements include confidentiality of data,user and device authentication,access control,and trust assurance among the things.However,recent reported events related to security attacks show colossal vulnerabilities among IoT devices capable of bringing security risks to the whole environment.One of the common uses of these devices by the attackers is to generate powerful distributed denial of service(DDoS)attacks.It is one of the most prominent attacking behaviors over a network by a group of geographically distributed zombie computers that interrupt and block legitimate users to use the network resources and hence,requires great attention.In this regard,the current work being novel in the field puts concentration on variants of DDoS attacks and their impact on IoT networks along with some of the existing countermeasures to defend against these attacks.The paper also discusses the detailed working mechanism of these attacks and highlights some of the commonly used tools that are deployed in such attack scenarios.

In this study,we investigate the relay selection (RS) problem in full-duplex energy-harvesting (FDEH) relay networks,where the relays are wirelessly powered by harvesting a portion of the received signal power from the source.We extend the investigation of the relay selection problem in FDEH relay networks to enable multiple relays to be selected simultaneously for improved performance.This is in contrast with existing studies on RS in similar setups,where only one relay can be selected in a transmission cycle.Our simulations show that selecting only a single relay is not always optimal,especially at low signal-to-noise ratios (SNRs).Furthermore,in this paper,we present the design of a greedy RS method with quadratic complexity for FDEH relay networks.Compared with the exhaustive-search-based RS,the proposed greedy RS achieves near-optimum performance in terms of the end-to-end capacity with significantly reduced complexity.

Although the worldwide demand for energy is increasing rapidly,the availability of current traditional resources (such as atomic and thermal power) is insufficient.Hence,artificial(man-made)energy sources are required.Advancements in wireless technology have raised the radio frequency (RF) levels in the environment.These RF waves are available at all times,unlike natural resources such as solar,hydro,and wind energy.The energy requirement and consumption of electronic devices have also been reduced over last decade.Hence,harvesting energy from RF waves and driving low-power devices are the simplest solutions.In the current work,a simplified monopole antenna,called a tri-stepped rectangular antenna for RF harvesting,is proposed.Here,the regular rectangular structure is modified to a step-like structure to achieve impedance matching and maximize the omni-directional gain at all mobile frequencies.This antenna functions with the LTE B5(850),GSM 900,GSM 1800,3G,4G,and ISM(2.4 GHz)systems.We have also demonstrated a proof-of-concept of energy harvesting using RF waves generated by mobile towers and Wi-Fi devices.The system can generate up to 12 mV(14.4 mW) and could charge a battery of rating 3.7 V,500 mAh.We anticipate this harvested energy to be used in driving the WSN node,Bluetooth devices,and mobile charging.

The tremendous and consistent increase in the volume and heterogeneity of traffic has resulted in major innovations in the telecommunication networks.In regard to the optical networks,existing studies have shown that by adopting a mixed line rate(MLR)strategy,the wavelength division multiplexed optical networks can cost-effectively respond to the diverse variety of traffic requirements which have heterogeneous service demands.Unlike existing studies which focus on various MLR network issues by considering deployment of the standard single mode fiber only within the network,in the current work,we investigate the signal quality deterioration due to the combined effects of dominant physical layer impairments for an MLR optical network conforming to the various ITU-T compliant fibers and also considering the optical frequency grid based on ITU-T Recommendation G.692.The main aim of our current study is to identify,for a given fiber,the modulation format configuration which provides the highest performance.We conduct extensive simulations on the considered MLR system using the obtained optimum channel spacing values between the single and mixed line rates.Our results show that the existence of 10 Gbit/s line rate has a detrimental effect on the 40 Gbit/s and/or 100 Gbit/s line rate; however,the 40 Gbit/s and/or 100 Gbit/s line rate’s effect on a 10 Gbit/s line rate is not so detrimental,as well as between the similar line rates.Overall,our results clearly show that choice of the line rate of both,the central channel and its adjacent channels,has a major effect on the MLR network performance.