Prediction-based dynamic load balancing method in 5G Internet of Vehicles CoMP (Coordinated Multi-Point) scene

Abstract

The invention relates to the technical field of 5G Internet of Vehicles, in particular to a prediction-based dynamic load balancing method in an Internet of Vehicles CoMP scene. The method comprises the following steps: counting the initial load condition of each base station, and taking a light-load base station as a cooperative base station of a full-load or overload base station; for a base station in an overload state, users are shunted firstly, part of edge users are shunted to an adjacent light-load base station for service, and then position distribution of the users is analyzed; for base stations in a full-load state or a light-load state, directly analyzing position distribution of users; the center user is directly served by the base station, and the edge users need to predict firstly and selectively adopt the CoMP technology to provide services for the users. Simulation analysis shows that the overload rate of the base station is obviously reduced in the non-peak period and the peak period, the rate of the user is also improved to a certain extent, the purpose of load balancing among the base stations is achieved, meanwhile, the communication quality of the user is guaranteed, and the life of the user is facilitated.

Description

technical field [0001] The present invention relates to the technical field of 5G Internet of Vehicles, in particular to a prediction-based dynamic load balancing method in a CoMP scenario of the Internet of Vehicles. Background technique [0002] With the rapid increase in the number of cars, the "tidal effect" has become more and more obvious in the driving time of vehicles, and the phenomenon of urban vehicle congestion has gradually become a social criticism, resulting in problems such as network freezes in the Internet of Vehicles and poor communication quality. It has brought some troubles to users. 5G and future wireless communication networks have the characteristics of dense distribution of communication users. With the increase of business types and the randomness of the distribution of mobile users in time and space, it is easy to cause some base stations to be overloaded. Not fully utilized, therefore, load balancing is particularly important. [0003] In the r...

AI Technical Summary

Problems solved by technology

[0002] With the rapid increase in the number of cars, the "tidal effect" has become more and more obvious in the driving time of vehicles, and the phenomenon of urban vehicle congestion has gradually become a social criticism, resulting in problems such as network freezes in the Internet of Vehicles and poor communication quality. Bring some troubles to users

5G and future wireless communication networks have the characteristics of dense distribution of communication users. With the increase of business types and the randomness of the distribution of mobile users in time and space, it is easy to cause some base stations to be overloaded. are not fully utilized, therefore, load balancing is particularly important

[0008] These algorithms are implemented under the CoMP technology, but these are not load-balanced for the Internet of Vehicles scene, and cannot better solve the load balancing problem proposed by the present invention for the CoMP scene of the Internet of Vehicles. Some existing load balancing algorithms based on prediction , such as the paper [25] (Luis Miguel Campos, Isaac D. Scherson. Rate of change loadbalancing in distributed and parallel systems. Parallel Computing, 26(2000), pp.1213-1230) uses the load change rate of the processor on the processor For dynamic load forecasting, although this algorithm is simple, the accuracy of this forecasting algorithm will be reduced for processor loads that change rapidly, are highly random, and do not have a stable and continuous state.

Images