Syslog is a critical data source for analyzing system problems. Converting unstructured log entries into structured log data is necessary for effective log analysis. However, existing log parsing methods demonstrate promising accuracy on limited datasets, but their generalizability and precision are uncertain when applied to diverse log data. Enhancements in these areas are necessary. This paper proposes an online log parsing method called DLLog, which is based on deep learning and has the longest common subsequence. DLLog utilizes the GRU neural network to mine template words and applies the longest common subsequence to parse log entries in real-time. In the offline stage, DLLog combines multiple log features to accurately extract the template words, creating a log template set to assist online log parsing. In the online stage, DLLog parses log entries by calculating the matching degree between the real-time log entry and the log template in the log template set. This method also supports the incremental update of the log template set to handle new log entries generated by systems. We summarized the previous works and validated DLLog using real log data collected from 16 systems. The results demonstrate that DLLog achieves high parsing accuracy, universality, and adaptability.

The proliferation of electric vehicles (EVs) couples the operations of traffic systems and power systems, necessitating the interdependent traffic-power modeling to optimize the on-road EV charging decisions. In this paper, the multi-period charging service pricing interactions between multiple charging network operators (CNOs) are discussed considering the dynamics of traffic-power systems. A multi-period user equilibrium model considering mixed vehicle types is formulated to describe the possible vehicle trip transitions between different periods in urban transportation network. For the power distribution network, a convexified multi-period AC optimal power flow model is adopted for local electricity market clearing. With the dynamics of traffic-power systems, the pricing interaction between multiple CNOs is a non-cooperative game. We analyze the existence of Nash Equilibrium with fixed-point theory, and propose an iterative method based on the best response strategy to find$\epsilon $-Nash Equilibrium. Numerical studies based on an interdependent power-traffic system consisting of two 18-node power distribution networks and one revised Nguyen-Dupuis transportation network demonstrate that with the proposed CNOs’ charging pricing strategy, the CNOs’ service profits can increase by 1%-3% while the PDNs’ operation costs decrease about 1%-2% for the interdependent traffic-power system.

Overlapping community detection is essential for revealing the hidden structure of complex networks. In this work, we present an overlapping community detection algorithm that selects community centers adaptively based on density peaks. The proposed algorithm, called the density-peak-based overlapping community detection (DPOCD) algorithm, defines point link strength and edge link strength to construct distance matrix. Unlike the density peaks clustering algorithm, by which cluster centers are selected manually, the DPOCD algorithm uses the linear fitting method to select community centers. To evaluate the feasibility of the presented algorithm, we compared it with other advanced methods on artificial synthetic network and real complex network datasets. The experimental results demonstrate that our method achieves excellent performance in large-scale complex networks and the robustness of the algorithm.

To address the communication and ranging challenges caused by underwater environment, we design dual modal devices for autonomous underwater vehicles (AUVs). The dual-modal design builds upon a co-axial ultrasonic and green laser beams which leverage different signal diverging patterns and different responses in the underwater environment by each modality to achieve robust adaptability. Here we report our recent progress in improving scanning and aiming capabilities for dual-modal beam steering. The core part is our Two-Axis Water-immersible Micro Scanning Mirror (WIMSM). We improve hinge design of WIMSM for larger scanning range. We incorporate high speed Hall effect sensor-based pose feedback channel to enable closed-loop scanning and aiming control. We design ultrasonic-assisted laser handshaking method to help AUVs to acquire optical underwater communication. We have prototyped our devices and tested them in a water tank. The initial results are promising.

Adopting extreme fast charging for electric vehicles will significantly reduce the charging time for electric vehicle owners, which will improve the public acceptance of electric vehicle. However, under the conditions of wide spread fast charging stations, large charging power of fast charging stations will bring nonnegligible impacts to the power system. For an aggregator that owns multiple fast charging stations, installing battery storage systems within the fast charging stations can reduce the impacts and give more flexibility. It will bring extra benefit if the aggregator participates in electricity markets by utilizing the flexibility of the storage. In order to deal with the operation and market participation problem for EV fast charging stations, this paper proposes bidding strategies in both energy and reserve markets for an aggregator of multiple fast charging stations with energy storage systems. Conditional Value at Risk (CVaR) based mixed integer quadratic programming formulation is built to hedge the risks of random charging demands and volatile market prices. The proposed formulation is converted into a linear programming problem and an iterative solution method is designed to reduce the computation time. Case studies based on the trajectory data of taxis in Beijing have been carried out. Simulation results show that the aggregator can achieve higher economic benefits while satisfying the fast charging demand of electric vehicles.

In the past decades, various approaches have been proposed to address the time series prediction problem, among which nonlinear autoregressive exogenous (NARX) models achieve great progresses in one-step time prediction. Although NARX models are capable of capturing long-term dependence of the time series data, the impact of associated attributes lacks enough attention. To cope with this issue, in this paper we propose a Multi-Attention algorithm based Recurrent Neural Network (RNN) to perform multivariate time series forecasting. In the first stage, given a raw multivariate time series segment, we obtain both relevant encoder hidden state and encoder hidden state of the associated attribute by employing input-attention and self-attention respectively. In the second stage, we use temporal-convolution-attention neural network to process the encoder hidden states and capture long-range temporal patterns. Finally, extensive empirical studies tested with four real world datasets (NASDAQ100, SML2010, Gas Sensor Array Temperature Modulation and Air Quality) demonstrate the effectiveness and robustness of our proposed approach.

This paper reports the development of a compact optical-acoustic frontend module for underwater communication and ranging. The module is enabled by a new water-immersible MEMS scanning mirror (WIMSM). It is capable of transmitting, receiving and steering co-centered and co-directional laser and ultrasound beams under water. To monitor its rotating angle in real time, scan position sensors based on Hall effect have been integrated into the WIMSM. The angular alignment of the laser and ultrasound beams in both transmission and reception modes has been examined. The experimental results show that the laser and ultrasound beams can remain aligned with less than 2.1 degrees under envelope of pan and tilt rotations. This capability is critical for the continuing development of the new bi-modal communication and ranging underwater Vehicles (AUVs).

In order to exploit the potential capacity of 5G, the deployment of ultra-dense small cells is an approach that can dramatically increase the radio resource reuse factor and network capacity. However, network densification with a large number of small cells brings challenges due to increased network complexity, deployment cost and inter-cell interference. In this paper, a new 5G architecture with virtual small cells (VSCs), which are dynamically formed by grouping a number of user devices in close proximity and adapted according to traffic condition, is proposed to improve the cost and energy efficiency compared with the traditional fixed deployment of small cells. In each virtual small cell, one mobile device is selected as a cell head (CH) to aggregate intra- cell traffic using unlicensed band transmissions and then communicates with its macro-cell base station in a licensed band through beamformed transmission, which reduces the inter-cell interference and improves spectrum efficiency. In this paper, a highly directional beamforming technique is employed to enable a dedicated inband fronthaul link for VSC. Our work focuses on how to design adaptive beamforming to minimize the transmit power under throughput requirements and power constraints. Both the mathematical analysis and simulation results demonstrate that VSCs can increase power efficiency dramatically while providing flexibility and reduced cellular load, when compared with macrocell only deployment and traditional fixed small cells scenario.

In order to support dramatically increased traffic from diverse network services, deployment of ultra dense networks to improve the overall capacity of the fifth generation (5G) wireless networks becomes inevitable. However, network densification with increased number of small cells brings significant challenges in terms of quality of service provisioning and deployment cost due to increased network complexity, signalling overhead and inter-cell interference. In this paper, virtual small cell (VSC), which is formed adaptively according to traffic condition and service requirements, is investigated as a solution for cost-effective and reliable service provisioning in 5G ultra dense networks. A K-means clustering based VSC formation scheme is proposed in this paper, and the corresponding protocol conversion for data transmission across unlicensed and licensed networks at cell head (CH) is developed. Based on the VSC architecture design, a new resource allocation algorithm is also proposed for VSC scenario in order to improve the system throughput with comparable fairness.

Future fifth generation (5G) wireless infrastructure tends to be highly heterogeneous, with dense small cells deployed overlay to cellular networks. Along with extremely high capacity and stringent latency requirements, security provisioning is becoming challenging in 5G Heterogeneous Networks (HetNets). Security key management could be difficult in small cells where users join and leave frequently, not to mention the limited capability of simplified access points (APs). On the other hand, frequent handovers and authentications in small cells also introduce unnecessary latency. Therefore in this article, we propose a software defined networking (SDN) enabled fast authentication scheme using weighted secure-context-information (SCI) transfer in order to improve authentication efficiency during handover and meet 5G latency requirement. The proposed algorithm is then applied in Neyman Pearson (NP) hypothesis test to authenticate users, which shows enhanced authentication accuracy and reduced latency in MATLAB simulations. Furthermore, we first analyze the SDN structure using priority queuing theory, and prove the performance of SDN enabled authentication handover.

Nowadays, self-driving vehicles which would shoulder the burden of driving and set free human on board are gradually becoming a reality. Consequently, the supporting of growing in-vehicle data traffic will be challenging in future 5G and vehicular networks, due to the high mobility nature of vehicles and the densified irregular distribution on road especially during rush time. Therefore in this paper, a Software-Defined Networking (SDN) enabled integrated 5G-VANET architecture is proposed to improve heterogeneous network (HetNet) management and aggregate vehicle traffic through IEEE 802.11p; a novel vehicle clustering method and dual cluster head design are then introduced to reduce signaling overhead and enhance the overall communication quality in 5G-VANET HetNet under the coordination of SDN. It is also proved by simulation that the proposed design reduced 5G users' blocking probability to the operators services with a back-up cluster head (CH) in each cluster, and also realized adaptive clustering without excessive SDN's processing delay.

Nowadays, data traffic in-car communication is increasing dramatically, due to the emerging technology of self-driving and on-board infotainment applications. The direct connections between vehicles and cellular infrastructures will introduce significant signalling overhead and excessive energy consumption, especially for congested and fast moving traffic. In order to improve energy efficiency and achieve green networking, an heterogeneous network, 5G-Vehicular Ad Hoc Network (5GVANET) is presented in this paper, which coupling the high data rates of VANET and the wide coverage area of 5G. In this integrated architecture, vehicles are clustered accordingly, and one vehicle in each cluster is selected as a gateway to support aggregated traffic. To ensure the capacity of the trunk link between the gateway and base station, a Non- orthogonal Multiplexed Modulation (NOMM) scheme is proposed in this paper to effectively aggregate the Vehicle-to-Infrastructure (V2I) traffic and further improve energy efficiency. NOMM splits data stream of each user into multi-layers and modulate them simultaneously. Sparse spreading code is also applied in partially superposing the modulated symbols on several resource blocks. Furthermore, we analyzed the energy efficiency of proposed NOMM scheme and traditional M-QAM theoretically. It was also validated by simulation results that NOMM provides less power consumption than M-QAM modulation.

The proliferation of mobile services and the explosive growth of data traffic has created new challenges in cellular networks. Mobile data offloading has attracted significant attention, since it has the ability to alleviate cellular burden by using complementary resources and thus, offers better services to end users. In this paper, we introduce intelligence into heterogeneous network management and propose a Software-Defined Networking based module framework, which includes Wi-Fi based partial data offloading and load balancing. Our objective is to make real time decisions for selectively offloading traffic and balancing loads, while taking network conditions and quality of service (QoS) into consideration. The proposed mechanisms are subject to system-level simulations which shows an improvement in load balancing, in terms of equilibrium extent and network stability. We also prove that with the proposed Wi-Fi partial data offloading algorithm, quality of service can be satisfied, while saving a significant amount of cellular resources through smart resource allocation.

To meet the requirement of very-high-rate-data transmission over wide bandwidths, carrier aggregation (CA) has been regarded as an important technology for Long Term Evolution-Advanced (LTE-A). It considers both primary component carrier (PCell) and secondary component carrier (SCell) and many operations are based on the PCell. In this paper, the load balancing performance of dynamic SCell measurement period relaxing in the LTE-A system with CA has been discussed. Furthermore, it gives the result of SCell measurement period relaxing window under different user equipment speed which is both energy-saving and harmless to the system performance.

To meet the requirement of very-high-rate-data transmission over wide bandwidths, carrier aggregation (CA) has been regarded as an important technology for Long Term Evolution-Advanced (LTE-A). It considers both primary component carrier (PCell) and secondary component carrier (SCell) and many operations are based on the PCell. In this paper, the load balancing performance of dynamic SCell measurement period relaxing in the LTE-A system with CA has been discussed. Furthermore, it gives the result of SCell measurement period relaxing window under different user equipment speed.

Handover Parameter Optimization (HPO) and Load Balancing (LB) are two Self-Organization network (SON) aspects which aim at improving LTE system handover performance and user's satisfaction respectively. However, there is often counteraction between LB and HPO, because LB would increase the frequency of inter-cell handover and correspondingly increase the possibility of handover problems. Furthermore, most of the LB and HPO jointly optimization methods don't consider the network allowed maximum radio link failure (RLF) ratio, which would increase the possibility of call dropping although the cell loading is balanced. In this paper we introduce the network allowed maximum RLF ratio as a key indicator and a dynamic hysteresis-adjusting (DHA) method to harmonize the two aspects. Furthermore, we take the realistic network situations into account to obtain a more reliable result. The proposed method is evaluated by a series of system-level simulation which witnesses an improvement in handover performance and number of satisfied users in LTE networks.

High Speed Railway(HSR) scenario was currently agreed as the main scenario in the 3GPP Rel.11 study item, Mobile Relay for E-UTRA. Usually, communications of high-speed railway systems suffer from problems such as Doppler spread, radio condition abrupt change and handover failure. Mobile relay is a promising scheme to solve these problems, but the quantitative performance improvement to HSR has not been fully evaluated and analyzed. In this paper, a high speed scenario with mobile relay integrated is presented to analyze these issues for LTE Advanced system. The proposed mobile relay solution with group mobility is evaluated by a series of system simulation which witnesses an improvement in train user throughput as well as system throughput, and higher handover success ratio with a decrease in radio link failure ratio.