China
Africa
Explore chapters and articles related to this topic
Ubiquity
Published in Vivek Kale, Digital Transformation of Enterprise Architecture, 2019
There are two main types of basic interactions: Synchronous interaction: The interaction protocol consists of a flow of control of two messages, a request then a reply or response. The sender sends a request message to the specified receiver and waits for a reply to be received e.g., a client component makes a request to a server component and gets a response.Asynchronous interaction: The interaction protocol consists of single messages that have no control of flow, a sender sends a message to a receiver without knowing necessarily if the receivers will receive the message or if there will be a subsequent reply, e.g., an error message is generated but it is not clear if the error will be handled leading to a response message.
Intelligent Agents
Published in Satya Prakash Yadav, Dharmendra Prasad Mahato, Nguyen Thi Dieu Linh, Distributed Artificial Intelligence, 2020
Rashi Agarwal, Supriya Khaitan, Shashank Sahu
Agent-based computing has a novel of abstractions and required adaptive methodologies. It advocates the development of new tools to support an agent-based system. There are two views points of engineering about the agent: one with a strong artificial intelligence view in which proactivity of agent and intelligence of agent are addressed; and in the second view, the software engineering view, an agent is a reactive or proactive thread for its environment. A complex system could not be developed with one agent only. It has many agents that have different roles in performing tasks of the complex system. When multiagents are working in a system, an interaction protocol is needed for interaction between agents. Agents also need local context in which agents will live.
Knowledge and agent-based system for decentralised scheduling in manufacturing
Published in Cogent Engineering, 2019
Salman Saeidlou, Mozafar Saadat, Guiovanni D. Jules
For scheduling activities to be delegated, there are two main components required, namely interaction protocol and negotiation mechanism. The interaction protocol governs the structure and timing of the data exchange between agents; contract net protocol and modified ring protocol are useful examples (Jules, Saadat, & Saeidlou, 2015; Owliya, Saadat, Jules, Goharian, & Anane, 2013). The negotiation mechanisms can be categorized into market-based or threshold-based approaches. The market-based approach caters for agents with self-interested goals. Agents compete and are rewarded if they exhibit desirable system-wide behaviours. The threshold-based approach is based on the probability of an agent accepting a preferred type of task when some events take place. However, market-based negotiation, which is a direct negotiation mechanism, has problems with communication overhead due to the constant exchanges of bids and the processing of those bids (Goldingay & Van Mourik, 2013; Shen, 2002). Threshold-based mechanisms which are often associated with indirect negotiation mechanisms such as stigmergy and bio-inspired coordination, do not suffer from communication scalability issues. The knowledge would be in the form of pheromone-type traces in the case of the threshold-based mechanism; and in the case of the market-based mechanism, knowledge would come from agent bids. Agents would operate within a context bound by rules, implicit data and inferred data (Yılmaz & Erdur, 2012).