Atsushi Terauchi

Detection and diagnosis of Inter-AS routing anomalies by cooperative intelligent agents

Verifying whether the routing information originating from an AS is being correctly distributed throughout the Internet is important for stable inter-AS routing operation. However, the global behavior of routing information is difficult to understand because it changes spatially and temporally. Thus, rapid detection of inter-AS routing failures and diagnosis of their causes are also difficult. We have developed a multi-agent-based diagnostic system, ENCORE, to cope with these problems, and improved its functions (ENCORE-2) through our experience in applying the system to commercial ISPs. Cooperative actions among ENCORE-2 agents provide efficient methods for collecting, integrating, and analyzing routing information observed in multiple ASes to detect and diagnose anomalies that human operators have difficulty in handling. ENCORE-2 is also applied to the hijacked route problem, which is one of recent major inter-AS issues.

Trust-Based Cooperative Action Control in Multi-agent Systems for Network Management

A mechanism to control cooperative actions in multi-agent systems for network management systems that consists of agents managed by different network management authorities (NMAs), such as ISPs and universities, is presented. In spite of cooperative work among NMAs being indispensable for Internet management, NMAs do not always trust or coordinate with each other because their intentions or policies may conflict. Therefore, in actual cases, cooperative work among NMAs is controlled with regard to the trustworthiness of the NMAs. Cooperative actions with no control may cause security problems, such as unintentional disclosure of network information. When a multi-agent system is deployed for network management, cooperative actions between agents must also be controlled in the same way as in actual network management. To achieve such control, our mechanism focuses on the trustworthiness of NMAs and controls cooperative actions on the basis of two parameters: 1) agent attributes and 2) reliability values. These parameters are referred or exchanged among the agents on an on-demand basis, and on their basis,the agents determine the availability of information of concern. Experimental results suggest that a network management system combined with our mechanism appropriately controls cooperative actions. These results demonstrate the feasibility of our basic design.

Diagnosis of IP-Service Anomalies Based on BGP-Update Temporal Analysis

Diagnosing IP-service anomalies requires network operators to analyze routing and service-application behavior at the protocol level from multiple viewpoints. For analyzing anomalies whose effects can be observed as static inconsistency among BGP routing tables, the multi-agent-based diagnostic system called ENCORE has been developed and successfully applied to actual autonomous systems (ASes). This system can integrate BGP routing information observed at multiple ASes and verify whether the routing information is being correctly distributed throughout the Internet. However, actual IP-service failures sometimes require temporal analysis of routing updates to identify specific causes and, moreover, comprehensive analysis from the viewpoint of interaction with other protocols than BGP. Thus, we have integrated such functions in the ENCORE architecture and developed a comprehensive diagnostic system called IP-MIND. It can consider historical BGP routing updates and the behavior of other protocols in conjunction with BGP-information reachability analysis in order to identify specific causes. It also enables diagnosis of a class of anomalies that cause reachability failures while not creating any inconsistency among the current BGP routing tables at multiple ASes.

A dynamic flow control mechanism based on a hierarchical feedback model for data center networks

Many studies have focused on developing Transmission Control Protocol (TCP) flow control mechanisms to enable more effective use of bandwidth in data center networks. However, the studies cannot achieve total optimization of bandwidth use for data center networks. This is because they do not take into account the path design of TCP flows and the hierarchical structure in data center networks through which TCP flows. To address this issue, we propose a TCP flow control mechanism for effectively using bandwidth by dynamically controlling paths through which TCP flows. To achieve this dynamic control, a hierarchical feedback model is used to obtain an optimal path establishment policy. With the model, all network components exchange their resource states with each other when each network component establishes a TCP flow. This enables each network component to learn the best TCP flow path for effective bandwidth use. An evaluation process shows that the mechanism achieves more effective bandwidth use and more efficient flow control than the conventional TCP flow control.

IM-DB: Information retrieval system for interactive network-status analysis

To manage the changing conditions of the Internet and diagnose the cause of anomalies within it, network operators must obtain and analyze information from network devices such as routers and network monitoring tools. Of that information, route information exchanged by Border Gateway Protocol (BGP) is essential for understanding the Internet at an inter-autonomous-system (AS) level. However, the information is not easily deployed for actual network management because of the enormous amount of the information and the difficulties in manipulating it. To overcome these problems, we propose IM-DB, a network information retrieval system for the interactive analysis and diagnosis of the Internet. Our IM-DB is based on a relational database and can store large amount of network information. By using the IM-DB, network operators can easily search, integrate, and explore information they require. We developed the first version of our IM-DB for storing BGP routing information. The experimental results using this prototype demonstrate the feasibility of our IM-DB.

Lisp-based agent platform and applications for inter-domain network management

The Internet consists of several thousand interconnected autonomous systems (ASes). For enabling the autonomous network management that ensures stable access at the inter-AS level and flexible inter-domain routing control, it is important to know how the routing information originated from an AS spreads throughout the Internet and to control inter-AS routing information using feedback actions based on observed network status at multiple ASes. Because each AS is controlled by a single administrative authority based on that ASs own policy, a cooperative distributed problem solving is desirable. To cope with these requirements, we have proposed a multi-agent-based inter-AS diagnostic system called ENCORE and inter-AS routing adjustment system called AISLE / VR. They consist of a collection of intelligent agents that are located in multiple ASes and perform collective observation, analysis, and control. These systems are constructed on the agent platform that provides utility functions on distributed environments. For the purpose of flexibility for incremental design and modification through the development and application phases, we adopted Lisp as the base language and have constructed stable systems that can demonstrate the effectiveness of our basic design for autonomous inter-AS network management.