Browsing by Author "Wang, Lei"

Now showing 1 - 4 of 4
  • Thumbnail Image
    ItemOpen Access
  • Thumbnail Image
    ItemOpen Access
    Bandit-based Delay-Aware Service Function Chain Orchestration at the Edge
    (2021-04-21) Wang, Lei; Ghaderi, Majid; Krishnamurthy, Diwakar; Safavi-Naini, Rei
    Mobile Edge Computing (MEC) enables both cloud computing and edge computing for mobile users, providing them with intensive computing resources and proximity to the data sources. When combined with network function virtualization (NFV), MEC provides users with promising end-to-end latency and management for mobile applications that requires multiple computing resources. Such applications are often handled in a fashion of service function chain (SFC), which designates a sequence of virtual network functions (VNF) for users’ traffic to traverse in order to realize their network application. In order to provide the user a tolerated perceived latency for a SFC-based application, many existing works have taken aim at optimal system-wide placement for SFC in heterogeneous scenarios yet fewer works have studied user-managed placement. In this paper, we formulate the user-managed SFC placement in MEC as a contextual combinatorial multi-arm bandit (C2MAB) problem and proposed BandEdge, a bandit-based algorithm for online SFC placement on edge, which consider user’s mobility and service preference while jointly optimizing their perceived latency and service migration delay, and then propose an offline exact approach for the role of performance benchmark. To fit the SFC placement problem in a bandit framework, we model the nodes and links to be arms by viewing them as delays and selects them according to a strategy that balances exploration and exploitation. Finally, we evaluate the proposed algorithm in extensive simulation and Mininet-WiFi emulation experiments, numeric simulation results show that the proposed algorithm can achieve close-to-optimum performance and outperform the greedy learning algorithms by at least 50 percent in terms of scalability. We further validate the superior performance of our proposed method in Mininet-WiFi emulation under different environmental parameters.
  • Thumbnail Image
    ItemOpen Access
    Probing Nucleation Mechanisms of Gas Hydrates via Molecular Simulations
    (2023-11-02) Wang, Lei; Kusalik, Peter G; Marriott, Robtert A; Tieleman, Peter DP; Clarke, Matthew Alexander; Patey, Grenfell
    Gas hydrates are ice-like solids where guest species can be encapsulated in water cages. Gas hydrates have received considerable attention because of the vast natural methane hydrates and various promising hydrate-based technologies. Relevant to the applications, a fundamental molecular-level understanding of nucleation behavior and mechanisms is necessary to support ongoing developments associated with gas hydrates. With the primary goal to provide significant insights into the mechanistic details of gas hydrate formation at the molecular level, this thesis investigates the behavior of gas hydrate nucleation and growth using Molecular Dynamic (MD) simulations along with Markov State Models (MSMs). The investigation of the nucleation behavior of mixed CH4/H2S hydrates within bulk aqueous solutions reveals the roles of different guest species in the hydrate formation process and the impacts of guest compositions, temperature, and different physical setups on the hydrate nucleation behavior. CH4 species has been consistently observed enriched in the hydrate phase relative to the aqueous solution, while increasing H2S composition can significantly enhance the rate of nucleation and growth of the hydrate through enhancement of the solubility of guests in solution. The investigation of the nucleation mechanisms of gas hydrates in water nanodroplets with pure-guest and mixed-guest species reveals key factors affecting hydrate nucleation behavior, including guest species, guest compositions, size of the nanodroplet, and temperature. The effects of temperature and the size of the water nanodroplet on the location of the initial hydrate nucleus have been explored. Utilizing water nanodroplet systems, the possible origins of the different effects of temperature control schemes on the behavior of hydrate formation have been addressed. It is found that the finite size of the surroundings is the origin of the influence of the temperature control schemes on the hydrate nucleation rates. Novel structural analyses, based on applying MSMs to data from gas hydrate simulations, have been used to identify and characterize a general transition network describing hydrate formation. This transition network of cage formation and the corresponding thermodynamic behavior confirm and validate that hydrate nucleation is essentially an ordering-in-stages process, where the early-stage behavior leading to a hydrate nucleus is dominated by entropic aspects.
  • Thumbnail Image
    ItemOpen Access
    Re-emergence of canine Leishmania infantum infection in mountain areas of Beijing
    (2023-03-30) Liu, Gang; Wu, Yuanheng; Wang, Lei; Liu, Yang; Huang, Wei; Li, Yifan; Gao, Mengbo; Kastelic, John; Barkema, Herman W.; Xia, Zhaofei; Jin, Yipeng
    Abstract Canine Leishmaniasis (CanL) is an endemic infectious disease in China, causing visceral Leishmaniasis (VL) and resulting in important public health problem. However, in the last 3 y, endemic trends have changed considerably and spatial–temporal aggregation areas have shifted from northwestern to central China. Although Beijing was an endemic area for CanL in the last century, this disease has not been reported in Beijing since control programs were implemented in the 1950s. In the present study, PCR and immunochromatographic (ICT) were used to estimate prevalence of Leishmania infection in domestic dogs living in Beijing, a VL re -emergencearea. In total, 4420 canine blood samples were collected at vet clinics in 14 districts of Beijing. Overall prevalence (percentage of dogs seropositive and/or PCR positive) of CanL infection in Beijing was 1.22% (54/4420). However, prevalence of CanL in the western mountain areas was 4.68% (45/961), significantly higher than that (0.26%, 9/3459) of the plains. In addition, multilocus sequence typing (MLST) of seven enzyme-coding genes was used to examine phylogenetic relationships of CanL strains. Forty-one Leishmania infantum isolates were well separated from the other strains and divided into five major clades (A to E) by MLST analysis. All clades were closely related to strains from Sichuan Province and Gansu Province. A phylogenetic tree, based on the MLST, revealed that L. infantum in Beijing was genetically related to strains from western endemic of Mountain type VL in China. In conclusion, CanL has re-emerged in Beijing, and almost 5% of dogs living in Beijing’s mountain areas were infected with L. infantum. The phylogenetic tree based on MLST effectively distinguished species of Leishmania and reflected geographical origins. Because dogs are considered a natural reservoir, comprehensive control measures including surveillance, phylogenetic analyses and management should be implemented to mitigate or eliminate Leishmaniasis.