Qi Guo’s research while affiliated with Chinese Academy of Sciences and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (4)


PerLLM: Personalized Inference Scheduling with Edge-Cloud Collaboration for Diverse LLM Services
  • Preprint

May 2024

·

36 Reads

·

Yuanhao Yang

·

Chang Zhao

·

[...]

·

With the rapid growth in the number of large language model (LLM) users, it is difficult for bandwidth-constrained cloud servers to simultaneously process massive LLM services in real-time. Recently, edge-cloud infrastructures have been used to improve the processing efficiency of large-scale LLM services. However, the diversity of task requirements and the dynamics of resources pose great challenges to inference scheduling, leading to the wastage of many resources. In this paper, we present PerLLM, a personalized inference scheduling framework with edge-cloud collaboration designed for diverse LLM services. For the complexity of multiple constraints and the decision-making process of edge-cloud collaboration, we integrate the upper confidence bound algorithm based on the constraint satisfaction mechanism in PerLLM. For diverse LLM services, PerLLM can optimize service scheduling and resource allocation solutions within the edge-cloud infrastructure to meet processing time requirements while minimizing energy costs. Experimental results from different model deployments show that PerLLM can effectively meet the processing time requirements of personalized services. Compared to other methods, PerLLM achieves 2.2x, 2.1x, and 1.6x throughput and reduces the energy cost by more than 50%.



Visual E 2 C: AI-Driven Visual End-Edge-Cloud Architecture for 6G in Low-Carbon Smart Cities

June 2023

·

12 Reads

·

13 Citations

IEEE Wireless Communications

With the rapid development of 6G wireless communication technology, the emergence of rich multimedia data for massive devices will lead to greater intensive computations and energy consumption. However, the requirements from both green communication and international low-carbon strategy can be challenging. In this article, we first systematically analyze the key challenges from the perspective of 6G networks for low-carbon smart city development. Then we propose an AI-driven visual end-edge-cloud architecture (E 2 C), which extends upon the conventional design from the perspective of human-machine fusion and carbon emission optimization. We provide systematical analysis and intelligent computing methods for carbon emission in visual end-edge-cloud architecture. This architecture can enable the provision of E 2 C AI intelligence for 6G networks through hybrid hierarchical optimization mechanisms. Finally, the experimental results demonstrate that our proposed architecture has better performance in smart cities, achieving lower carbon emissions compared to traditional methods.


Citations (3)


... Approach Optimization objective SPINN [73], Adaptive offloading [190], Shoggoth [158], Sniper [94], JAVP [176], Auction-base [41] Distributed DNN inference over end devices, edge, and cloud. ...

Reference:

AI-Powered Urban Transportation Digital Twin: Methods and Applications
JAVP: Joint-Aware Video Processing with Edge-Cloud Collaboration for DNN Inference
  • Citing Conference Paper
  • October 2023

... Meanwhile, work is underway on 6G technology, which promises to further increase data speeds and connectivity [141][142][143]. This development will set a new standard for future smart cities and energy management solutions [144][145][146][147]. ...

Visual E 2 C: AI-Driven Visual End-Edge-Cloud Architecture for 6G in Low-Carbon Smart Cities
  • Citing Article
  • June 2023

IEEE Wireless Communications

... In Guo, Sun and Ji, (2022),An architecture for multi-tenant personalised warehousing modes, a platform for personalisation warehouse management in SaaS, and services for data isolation, interface customisation, fast positioning, and on-demand customisation are all proposed to solve the problem of traditional customisation not working in SaaS. The results of the experiments demonstrate that the suggested multi-tenant customised warehouse architecture may make advantage of shared storage resources, highly automated warehousing, and on-demand customisation for warehouse administration, data security, and user experience, in addition to isolating and customising data [37]. This Elgedawy, (2015),bound them closely to the selected data consistency methods, which hinders the adaptability and development of business services. ...

Design of Personalization Warehouse Management Platform Based on SaaS Model
  • Citing Conference Paper
  • May 2022