A Dynamic Routing Algorithm Based on Dual Estimators

Abstract

The invention discloses a dynamic routing selection method based on dual estimators, which has a data transmission system, a queue storage system, and a routing selection system at routing nodes, including the following steps: (1) acquiring environmental information; (2) setting Initial value; provide two estimators, the parameters of each estimator include current state, action, initialization value function and reward information; (3) In each time step, each routing node in the network simultaneously performs data transmission and The work of data reception; (4) the routing selection method during transmission is to obtain a random number, when the random number is greater than, select the optimal action according to the value function of the estimator, otherwise randomly select the action; (5) through the double estimator Update the value function; (6) Repeat steps (3) to (6) for each time step. The invention reduces the congestion of the network, can learn a better strategy at a lower cost, improves the performance of the network, and can effectively deal with the problem of route selection in a highly random network.

Description

technical field [0001] The invention relates to a network routing selection method, in particular to a dynamic routing selection method, especially a dynamic routing selection method suitable for high random networks. Background technique [0002] In recent years, with the development of network technology, the overall communication data of the network has been increasing. In an increasingly complex network, it is particularly important to formulate effective routing policies. Traditional static routing selection algorithms usually use fixed rules, such as routing tables, for routing selection. This method cannot make corresponding adjustments in time for network state fluctuations, so it is difficult to apply to large-scale networks with uncertain loads. The dynamic routing algorithm can automatically adjust the routing strategy according to the current network state, but the general dynamic algorithm requires global information and has high complexity, which increases th...

AI Technical Summary

Problems solved by technology

However, there are also problems if the Q-learning method is used directly in the routing algorithm

Since Q learning uses the maximized estimator to estimate the function, it will generate a maximization deviation, which will cause the value function learned by the routing algorithm directly using Q learning to be overestimated in the high random network, thus affecting the learning rate. This makes the performance of this routing algorithm poor in a highly random network environment

Images