Zhi Qi

Title: Zhi Qi and Deep Reinforcement Learning for Edge Cache Replacement.

Introduction

Zhi Qi is an inventor associated with Chongqing University of Posts and Telecommunications in Chongqing, China. His patent work addresses edge caching, a technique for storing content closer to users to reduce the time needed to access it.

Latest Patents

Qi is associated with 1 patent: “Methods for Edge Cache Replacement Based on Deep Reinforcement Learning.” The invention describes a method that models a system network and obtains the caching state of a mobile edge computing (MEC) server. It calculates the total delay users experience when retrieving cached content, then uses that measure to define an optimization objective and formulate a Markov decision process. A dynamic content-replacement algorithm based on deep reinforcement learning selects content for the MEC server’s cache. The method aims to minimize users’ average content-access delay while accounting for the server’s cache capacity.

Career Highlights

Qi’s patent reflects work at the intersection of edge computing, content delivery, and machine learning. By applying deep reinforcement learning to cache replacement, the method seeks to adapt caching decisions to network conditions and content-access needs.

Collaborations

The provided information identifies Qian Liu and Sihong Wang as Qi’s coworkers. Their specific contributions to the patent are not stated.

Conclusion

Zhi Qi’s patent presents a reinforcement-learning approach to managing MEC server caches, with the goal of reducing content-access delays for users.