40 results about "Queue stability" patented technology

An advance is made over the prior art in accordance with the principles of the present invention that is directed to a new approach for a system and method for a buffer management scheme. Certain embodiments of the invention improve the response of AQM schemes with controllable parameters to variations of the output rate of the bottleneck buffer. The impact on TCP performance can be substantial in most cases where the bottleneck rate is not guaranteed to be fixed. The new solution allows AQM schemes to achieve queue stability despite continuous variations of the bottleneck rate.

The invention relates to a wireless self-backhaul resource scheduling method based on system stability, which belongs to the field of mobile communication. This method guarantees the system queue stability, wireless backhaul constraints, and the rate matching corresponding to the system achievable rate and the minimum user service quality, that is, the overflow probability satisfies certain constraints, and is carried out with the goal of maximizing the relative user quality satisfaction level. Dynamic resource scheduling. In each discrete time slot, according to the user's channel state information and the queuing conditions jointly considered by the macro base station and the small base station, the appropriate wireless resources are allocated to each user, and the queue length is dynamically adjusted according to the overflow probability in order to change the actual reachability of the system. Rate and joint dynamic adjustment of the wireless self-backhaul small base station and the backhaul and access resource allocation ratio of each user. The wireless self-backhaul resource scheduling method based on system stability proposed by the present invention can maintain system queue stability while maximizing the relative user quality satisfaction level.

The invention relates to a Lyapunov-based virtual resource allocation method for network utility maximization, which belongs to the technical field of mobile communication. This method uses Lagrangian duality and KKT conditions to calculate the frequency spectrum, computing resources, and buffer resources of the base station occupied by each user; in each time slot, update the queue status information of the user; perform virtual resource allocation and system stability Joint optimization to achieve the purpose of maximizing network utility. The invention aims at the allocation of virtual resources in the wireless network, and simultaneously optimizes the allocation of virtual resources and the stability of the system in order to ensure the stability of the system. Taking wireless spectrum resources as income, and backhaul link resources, computing resources, and cache resources as expenses, and considering the stability of system queues and the differences of users of different virtual service providers, the purpose of maximizing utility is finally achieved. This method can maximize the average total revenue of the wireless virtual network while ensuring the stability of the system queue.

The invention provides a sliding model control method for a desired safety margin car-following model so as to avoid the influence of system disturbance errors caused by the uncertain traffic environment and a vehicle system on the traffic flow stability. According to the technical scheme, under the condition that system disturbance is considered, the sliding model control method is introduced, and a novel DSM car-following model is established. Compared with an original DSM car-following model, the provided sliding model control method can effectively improve the queue stability, and technical support is provided for stability control methods of the fields of self-adaption cruise control, automatic driving and the like of vehicles.

The invention discloses a resource optimization method and electronic equipment in mobile edge computing task offloading. The method includes: constructing a queue stability index aiming at minimizing the length of the user terminal task queue and the edge server task queue length; constructing network resource overhead Index; use the queue stability index as a constraint to construct a random network resource optimization model; introduce Lagrangian multipliers to relax the constraints of the stochastic network resource optimization model to build a coupled model; solve it based on the momentum stochastic gradient descent algorithm By coupling the model, the optimal resource allocation decision for each time slot is obtained. The invention constructs a dual model of a random network resource optimization model based on the original-dual coupling theory, and solves the dual problem online based on the momentum stochastic gradient descent algorithm, accelerates the convergence of the algorithm and reduces the backlog of task queues, compared with the existing resource optimization method , can further reduce the task queue backlog without increasing network resource overhead.

The invention relates to a virtual resource allocation method based on a self-backhauled small cell network, and belongs to the technical field of mobile communication. A virtual resource allocation method based on a self-backhauled small cell network, which obtains available virtual resource information, user status information, and slice requirement information of each base station through a virtual network manager, and the user status information includes user channel quality information and buffer queues Information, and then allocate virtual resources for users served by each service provider, and dynamically adjust the frequency band ratio of the self-backhaul small base station for wireless backhaul; at the same time, in the allocation process, the minimum average rate of each service provider is guaranteed requirements, system stability, and small cell backhaul capacity constraints. The virtual resource allocation method provided by the invention achieves the purpose of making full use of frequency band resources, and at the same time reduces the user's processing time delay at the small base station, and can maintain the stability of the system queue while increasing the average total revenue of the service provider.

The present invention relates to the field of resource optimization in the Internet of Vehicles network, in particular to a resource optimization method for the Internet of Vehicles network based on non-orthogonal multiple access technology. When processing vehicle tasks in the vehicle edge computing system assisted by NOMA, the vehicle edge computing The principle of minimizing the total energy consumption of the system is to determine the system's unloading and caching decisions, and the allocation of computing and caching resources, that is, considering the random traffic arrival of vehicle users and queue stability, computing unloading decisions and content caching decisions through joint optimization, and The allocation of computing and cache resources is defined as a stochastic optimization problem; using Lyapunov optimization theory, a dynamic joint computing offloading, content caching, and resource allocation algorithm is proposed to solve this problem, and it is decoupled into two independent sub-problems. Using 0-1 integer programming and linear programming to solve two sub-problems; the invention can effectively handle computing resources of mobile edge computing servers and reduce system energy consumption.