Mobile Agent Building Block of Distributed Systems

It is an abstraction. This article is a new technology for mobile agents that makes it easy to implement, design and maintain distributed systems. In this article I will explain the advantages of mobile agents and explain their impact on distributed system design. Mobile agents provide an effective way to reduce network traffic and overcome network latency. Perhaps most important is the possibility of asynchronous and autonomous management of the process by which the mobile agent is created.

What I am doing here is to examine the idea of ​​distributed computing and see if the basic idea of ​​loose coupling or low information coupling can be used as an inspiration for how to build a multi-proxy DL system It is to do. The obstacle of the main concept is the concept of meta level information that exists in a distributed system. However, there are still many common ideas that should be clearly used. There are two aspects to this exploration aspect: one is to decide the basic principle and the other is to decide constraints on multi-agent system communication. Constraint identification, or discourse language definition, allows you to learn to provide limited space for exploration and therefore to adjust.

In the inevitable transition to a more modular multipurpose multiagent deep learning system, we need to start exploring the same loose coupling principle as supporting distributed system coordination. An important criterion for building an effective DL system is better generalization. Generalization means a lot, but you can accept at least an intuitive explanation. In other words, generalization means a highly adaptable system. In the distributed architecture, adaptability is improved by the principle of loose coupling, so valuable ideas on the best way to build a multi-agent architecture like DL can be gained.

Group communication is an important component for many intrusive and fault tolerant distributed systems. The group communication system (GCS) was developed to ensure the consistency of the state between the processes that make up the group when a failure occurs. Although a lot of work has been done to endure the GCS with benign failure (see a comprehensive survey), the problem of establishing a GCS that can tolerate malicious corruption of a particular group member has recently gained momentum It is increasing.