Yipeng Wang

Yipeng Wang

Auto-generated profile, based on publicly available information

Yipeng Wang

Intel Corporation

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 November 2017 in IEEE Communications Letters
Verified
Journal Article
Closed access
SLShuaijun Liu
Shuaijun Liu
XHXin Hu
Xin Hu
YWYipeng Wang
Yipeng Wang
Gaofeng CuiGaofeng Cui
Gaofeng Cui
WWWeidong Wang
Weidong Wang
With the rapid development of low earth orbit (LEO) constellation satellite networks and on-board processing (OBP), in-network content caching becomes feasible which can further improve system performance. In order to minimize user terminals content access delay, a novel caching algorithm by optimizing content placement in LEO satellite constellation networks is proposed in this letter. The formulated problem is proved to be NP-Complete (NPC) by reducing it to the 2-Disjoint Set Cover (2DSC) problem which is a known NPC problem. This paper considers the interactions among distributed satellites for making individual content decisions. Then, this paper adopts an exchange-stable matching (ESM) algorithm based on many-to-many matching game with externalities. Simulation results show ESM algorithm can improve the system performance in terms of content access delay.

1 June 2020
Verified
Conference Paper
Closed access
BLBin Li
Bin Li
YWYipeng Wang
Yipeng Wang
Ren WangRen Wang
Ren Wang
CTCharlie Tai
Charlie Tai
RIRavi Iyer
Ravi Iyer
KAKrste Asanovic
Krste Asanovic
Network function virtualization (NFV) technology attracts tremendous interests from telecommunication industry and data center operators, as it allows service providers to assign resource for Virtual Network Functions (VNFs) on demand, achieving better flexibility, programmability, and scalability. To improve server utilization, one popular practice is to deploy best effort (BE) workloads along with high priority (HP) VNFs when high priority VNF's resource usage is detected to be low. The key challenge of this deployment scheme is to dynamically balance the Service level objective (SLO) and the total cost of ownership (TCO) to optimize the data center efficiency under inherently fluctuating workloads. With the recent advancement in deep reinforcement learning, we conjecture that it has the potential to solve this challenge by adaptively adjusting resource allocation to reach the improved performance and higher server utilization. In this paper, we present a closed-loop automation system RLDRM 1 1 RLDRM: Reinforcement Learning Dynamic Resource Management to dynamically adjust Last Level Cache allocation between HP VNFs and BE workloads using deep reinforcement learning. The results demonstrate improved server utilization while maintaining required SLO for the HP VNFs.

19 November 2020
Verified
Book Chapter
Closed access
PDF file
YWYipeng Wang
Yipeng Wang
JSJasvinder Singh
Jasvinder Singh
ZTZhe Tao
Zhe Tao
LMLiang Ma
Liang Ma
HZHeqing Zhu
Heqing Zhu
86Running the network functions purely in software is disruptive; it increases flexibility and agility of the infrastructure with the aim of reducing the total cost of ownership (TCO). The idea has been proved by the leading cloud service providers who are transforming their internal networking deployment. However, the transformation journey is not smooth. Indeed, there are challenges in a rapid rollout of the software network functions (often using open source) in the production systems. The functional requirements are easy to be satisfied, but the performance requirement is more difficult to meet, and it takes nontrivial software expertise and skills to build a high-performance networking system. The packet forward mechanism such as run-to-completion (RTC) model, or pipeline model, is a good reference to build the real-world performant packet processing workloads on the multicore CPU architectures.