Embedded Files
What kind of research?
We are working on a paper that describes how to efficiently process large amounts of data sent from cameras and sensors using nearby computers (edge computing). Data includes “inference tasks” that require immediate answers and “learning tasks” that make AI smarter. We reduce waiting times by adjusting the order and location of processing according to the characteristics of each task.
Additionally, we are developing mechanisms to analyze network traffic and distribute loads effectively to prevent congestion, ensuring smooth overall operation. This will enable stable use of AI and communication services throughout the city.
Load balancing for edge computing
Load balancing for edge computing
In edge computing, appropriate load balancing can significantly reduce processing time. Our laboratory proposes appropriate load balancing methods based on various tasks that arrive at edge servers.
Delay Analysis by queueing theory
Delay Analysis by queueing theory
The characteristics of latency are extremely important factors in the design of load balancing methods for edge computing and in the analysis of communication protocols for special environments such as outer space. In our laboratory, we analyze the latency characteristics of communications theoretically and experimentally while identifying factors that affect latency in various network environments.
Page updated
Google Sites
Report abuse