Method for self-adaptive load balancing based on future load prediction

Abstract

The invention provides a method for self-adaptive load balancing based on future load prediction. According to the method, the Markov chain is used for predicting the load of a network at the next moment, so that the threshold of a trigger load balancing method is adjusted in a self-adaptive mode and modeling of an access control method is conducted. According to the method for self-adaptive load balancing, the probability of the light load or the heavy load in a future moment is calculated according to the previous load condition of the network through the transition probability defined through the method; a future load benefit value of the network is calculated through the calculated probability according to a load benefit function defined through the method. When a user requests access to be switched or an access request is newly sent in a network, the network with the small load benefit value is preferably selected to serve as a target network for access, so that the load of the whole heterogeneous network is balanced and the call drop rate and the access blocking rate of the switch are effectively reduced. Meanwhile, if it is predicated that the probability of the future light load is large, the threshold of the trigger load balancing method is dynamically improved, and the network is prevented from executing unnecessary load balancing operation.

Description

technical field [0001] The invention relates to the technical field of wireless resource allocation in heterogeneous networks, in particular to an adaptive load balancing method based on future load prediction in a wireless heterogeneous network. Background technique [0002] The development direction of the future communication system is the integration of heterogeneous networks. In order to better realize the integration of heterogeneous networks, it is necessary to study the wireless resource management technology applicable to the heterogeneous network environment, so as to achieve overall planning and optimization of the use of wireless resources of each access network. 1. The purpose of improving the wireless resource utilization rate of the heterogeneous system. As an important aspect of wireless resource management, load balancing technology can effectively avoid uneven system load distribution and expand system capacity, which is the focus of current research. [0...

AI Technical Summary

Problems solved by technology

[0006] The disadvantage of this method: the load balancing method can only be executed after the network overload reaches the previously set threshold, there is no predictive and preventive function, and during this period, if many users request vertical switching access, the overload at this time Whether the network accepts or rejects the access request. If it is accepted, the performance provided by the network will be reduced and the Qos of other users will be affected. If it is rejected, the call blocking rate or call drop rate will increase.

The setting of the threshold is not dynamically adjusted according to the actual situation of the future load of the network

Images