arXiv Science⌕ Search

arXiv subjects

Tongtong Lin

Publications and source records attributed to Tongtong Lin.

2 recordsLinked to original sources

Superregular Breathers in the Sasa--Satsuma Equation: Spectral Conditions and Modulation Instability

In this work, we study the superregular breathers (SRBs) of the Sasa--Satsuma (SS) equation by combining a Darboux construction with a linear stability calculation. The reduction symmetry of the scalar SS Lax pair makes the fundamental SRB construction degenerate. The paired quasi-Akhmediev branches nevertheless yield second-order SRBs in the focusing sector $n|_{γ=0}<0$; neither the defocusing sector nor the focusing sector $n|_{γ=0}\geq0$ supports this construction. Linearization about the same plane-wave background gives the modulation-instability gain and, on the admissible branches, reproduces its relation to the group-velocity difference. The carrier wavenumber $b$ enters through the background rather than as a coefficient of the SS equation, and changes the spectral roots and instability interval. In the formal NLS limit, the result reduces to the familiar NLS gain. Numerical evolution of finite-width periodic perturbations produces the predicted second-order profile over the displayed propagation interval.

nlin.SI↗

Data-Driven Online Resource Allocation for User Experience Improvement in Mobile Edge Clouds

As the cloud is pushed to the edge of the network, resource allocation for user experience improvement in mobile edge clouds (MEC) is increasingly important and faces multiple challenges. This paper studies quality of experience (QoE)-oriented resource allocation in MEC while considering user diversity, limited resources, and the complex relationship between allocated resources and user experience. We introduce a closed-loop online resource allocation (CORA) framework to tackle this problem. It learns the objective function of resource allocation from the historical dataset and updates the learned model using the online testing results. Due to the learned objective model is typically non-convex and challenging to solve in real-time, we leverage the Lyapunov optimization to decouple the long-term average constraint and apply the prime-dual method to solve this decoupled resource allocation problem. Thereafter, we put forth a data-driven optimal online queue resource allocation (OOQRA) algorithm and a data-driven robust OQRA (ROQRA) algorithm for homogenous and heterogeneous user cases, respectively. Moreover, we provide a rigorous convergence analysis for the OOQRA algorithm. We conduct extensive experiments to evaluate the proposed algorithms using the synthesis and YouTube datasets. Numerical results validate the theoretical analysis and demonstrate that the user complaint rate is reduced by up to 100% and 18% in the synthesis and YouTube datasets, respectively.

eess.SP↗