Flow forecast based periodical and adaptive convergence method and system

Abstract

The invention pertains to the optical communication technical field, which provides a periodic self-adaptive converging method and a system based on traffic prediction. The concrete method is that business volume reaching an converging queue within self-adaptive period finished at present is received and predicted value of the business volume reaching the converging queue next period is calculated; the length of the converging queue at present is read and estimated emergent demand of the next self-adaptive period is obtained; threshold is determined according to preset emergent demand and length threshold value of the converging queue used in next self-adaptive period is predicted; when the length of the converging queue reaches the estimated value of the length threshold, or no converging happens during the present period, the converging queue is triggered once. The invention promotes self-adaptive precision of converging and increases flexibility of converging, and the most suitableself-adapting granularity can be selected according to concrete demand, thus maximizing cost performance of the product.

Description

technical field [0001] The present invention relates to the technical field of optical communication, in particular to a method and system for periodic self-adaptive convergence based on flow prediction. Background technique [0002] OBS (Optical Burst Switching, Optical Burst Switching) is a switching technology whose switching granularity is between OCS (Optical Circuit Switching, Optical Circuit Switching) and OPS (Optical Packet Switching, Optical Packet Switching). It is more flexible and efficient than OCS, can well support bursty packet services, and at the same time reduces the requirements for optical devices, and is an effective solution for next-generation optical networks. The core idea of ​​OBS is to separate the control information from the data information, so as to make full use of the huge bandwidth resource of optical fiber and the flexibility of electronic control. Data is transmitted and exchanged in the optical domain in the form of bursts, and the BHP ...

AI Technical Summary

Problems solved by technology

[0004] However, the aggregation method based on the adaptive length threshold has weak adaptive ability

It qualitatively estimates the trend of traffic changes by comparing the relationship between the actual burst length and the length threshold window. This simple prediction mechanism cannot provide more accurate traffic size information, so the aggregation method based on this prediction mechanism is also difficult to implement. The optimal dynamic adjustment of the length threshold, that is, although the method is adaptive to a certain extent, its adaptive accuracy is very limited; and the aggregation method based on the adaptive length threshold always uses a single burst as the basic unit. Adaptive adjustment of the threshold, its adaptive granularity is fixed, lack of flexibility

Images