Yousuke Takahashi

Traffic engineering based on stochastic model predictive control for uncertain traffic change

Traffic engineering (TE) plays an essential role in deciding routes that effectively use network resources. This is particularly important when one considers the increasing time variation of Internet traffic such as streaming and cloud services. Traffic engineering with traffic prediction is one approach to stably accommodating time-varying traffic. This approach calculates routes from predicted traffic to avoid congestion, but predictions may include errors that instead cause congestion. We propose a prediction-based traffic engineering method that is robust to prediction errors by considering the probability distribution of predicted traffic. Our approach is based on a control-theoretic approach called stochastic model predictive control. Routes are calculated using a probability distribution of prediction errors so that the occurrence probability of congestion is lower than an operator-specified level. By considering the multi-step future dynamics of traffic, the routes are changed gradually to avoid route oscillation. We also show a relaxation method for unreliable far-future probabilistic constraints to avoid overly conservative route changes. Through simulations using backbone network traffic traces, we demonstrate that our method can accommodate most traffic variations under a given target link capacity without sudden large routes changes.

Traffic prediction for dynamic traffic engineering

Traffic engineering with traffic prediction is a promising approach to accommodate time-varying traffic without frequent route changes. In this approach, the routes are decided so as to avoid congestion on the basis of the predicted traffic. However, if the range of variation including temporal traffic changes within the next control interval is not appropriately decided, the route cannot accommodate the shorter-term variation and congestion still occurs. To solve this problem, we propose a prediction procedure to consider the short-term and longer-term future traffic demands. Our method predicts the longer-term traffic variation from the monitored traffic data. We then take account of the short-term traffic variation in order to accommodate prediction uncertainty incurred by temporal traffic changes and prediction errors. We use the standard deviation to estimate the range of short-term fluctuation. Through the simulation using actual traffic traces on a backbone network of Internet2, we show that traffic engineering using the traffic information predicted by our method can set up routes that accommodate traffic variation for several or more hours with efficient load balancing. As a result, we can reduce the required bandwidth by 18.9% using SARIMA with trend component compared with that of the existing traffic engineering methods.

Flow aggregation for traffic engineering

Although the use of software-defined networking (SDN) enables routes of packets to be controlled with finer granularity (down to the individual flow level) by using traffic engineering (TE) and thereby enables better balancing of the link loads, the corresponding increase in the number of states that need to be managed at routers and controller is problematic in large-scale networks. Aggregating flows into macro flows and assigning routes by macro flow should be an effective approach to solving this problem. However, when macro flows are constructed as TE targets, variations of traffic rates in each macro flow should be minimized to improve route stability. We propose two methods for generating macro flows: one is based on a greedy algorithm that minimizes the variation in rates, and the other clusters micro flows with similar traffic variation patterns into groups and optimizes the traffic ratio of extracted from each cluster to aggregate into each macro flow. Evaluation using traffic demand matrixes for 48 hours of Internet2 traffic demonstrated that the proposed methods can reduce the number of TE targets to about 1/50 ~ 1/400 without degrading the link-load balancing effect of TE.

Traffic prediction for dynamic traffic engineering considering traffic variation

Traffic engineering with traffic prediction is one approach to accommodate time-varying traffic without frequent route changes. In this approach, the routes are calculated so as to avoid congestion based on the predicted traffic. The accuracy of the traffic prediction however has large impacts on this approach. Especially, if the predicted traffic amount is significantly less than the actual traffic, the congestion may occur. In this paper, we propose the traffic prediction methods suitable to the traffic engineering. In our method, we perform preprocessing before the prediction in order to predict the periodical variation accurately. Moreover, we consider the confidence interval for the prediction error and the variation excluded by the preprocessing to avoid the congestion caused by the temporal traffic variation. In this paper, we discuss three preprocessing approaches; the trend component, the lowpass filter, and the envelope. Through simulation, we clarify that the preprocessing by the trend component or the lowpass filter increases the accuracy of the prediction. In addition, considering the confidence interval achieves the lower link utilization within a fixed control period.

Optimizing Cache Location and Route on CDN Using Model Predictive Control

In content delivery services, cache-server selection and route control are independently operated by content delivery network (CDN) providers and Internet service providers (ISPs), respectively. However, the number of ISPs providing CDN service has been increasing, and they can optimize the cache-server and delivery-route selection simultaneously. In this paper, we investigate the effect of jointly controlling these two operations. To continuously maintain a desirable state over a long time span in an environment in which estimating future demand is not easy, we should use an optimization method with which we can repeat the optimization procedure continuously. Therefore, we propose a method of optimizing the cache-server selection using the model predictive control (MPC), which has been widely used in system control. We also propose a method of simultaneously optimizing both the cache-server and delivery-route selection using the MPC. Through numerical evaluation using two actual network topologies of Tier-1 ISPs and the access log data of VoD services, we confirm that we can achieve almost the same effect by optimizing only cache-server selection using the MPC compared with optimizing both operations simultaneously. We also confirm that the number of failed requests that are not accommodated in the system can be reduced by about 1/30 with the proposed methods using the MPC.

Framework for traffic engineering under uncertain traffic information

Traffic engineering (TE) plays an essential role in deciding routes that effectively use network resources. The TE controller should handle the uncertainty due to the lags and lacks of collected network information. Many previous work partially tackled this uncertainty problem in various aspects e.g. data collection, estimation, prediction, routing with uncertain traffic. However there are few studies about integrating these partial processes to achieve the cooperation. In this paper, we proposed a framework to integrate partial processes in a whole TE process, and formulate it according to the Bayesian approach. In our framework, decision making process considers how the decision affects not only network but also other processes by modeling the behavior of the process as conditional probability. Thus, the cooperation of different processes is expected to be achieved.

Separating predictable and unpredictable flows via dynamic flow mining for effective traffic engineering

For Internet service providers to efficiently use network resources, they need to conduct traffic engineering to dynamically control traffic routes to accommodate traffic with limited network resources. The performance of traffic engineering depends on the accuracy of traffic prediction. However, the volume of network traffic has been changing drastically in recent years due to the growth of various types of network services, making traffic prediction increasingly difficult. Our simple ideas to overcome this challenge are to separate traffic into predictable and unpredictable parts and to apply different control policies to predictable and unpredictable traffic. To promote these ideas, we use software-defined networking technology, particularly Open-Flow, that can control macroflows defined by any combination of L2-L4 packet header information such as 5-tuple. In this paper, we therefore propose the macroflow-generating method for separating traffic into predictable macroflows that have little traffic variation and unpredictable macroflows that have large traffic variation within a limited flow table size. We also propose a macroflow-based traffic engineering scheme that uses different routing policies in accordance with traffic predictability. Simulation evaluation results suggest that our proposed scheme can reduce the maximum link load in a network at the most congested time by 34% and the average link load in a network on average by 11% compared with the current traffic engineering schemes.

Hierarchical traffic engineering based on model predictive control

Traffic engineering (TE) plays an essential role in deciding routes that effectively use network resources. Since managing the routes of a large network takes a large overhead, multiple controllers are introduced in the network which hierarchically decide the routes. We call this approach hierarchical TE. In hierarchical TE, avoiding the route oscillation is a main problem since routes change at a layer causes the additional routes changes at other layers. The existing hierarchical TE avoids these route oscillation by setting the longer control interval on the upper layer. This approach, however, causes another problem that the routes change of upper layer delays to traffic changes. In this paper, we propose a hierarchical TE method called hierarchical model predictive traffic engineering (hierarchical MP-TE) which avoids routing oscillation without setting the long control interval. In hierarchical MP-TE, each control server gradually changes the routes based on the traffic prediction to stabilize the routing instead of setting the long control interval. Through the simulation, we show that the hierarchical MP-TE achieves the routing convergence with the short control interval.