Projects in Huawei (09/2015-08/2017)
1. Controller Design and Implementation in Application-Driven Networks (ADN)
09/2015-08/2017
Position: Leader of the Theory Group, Future Network Theory Lab
Contributions:
Position: Leader of the Theory Group, Future Network Theory Lab
Contributions:
- a general model of modeling network as a feedback control system.
- a hybrid network architecture with fast and slow control planes.
- a QoS-aware traffic engineering algorithm for slow control plane. Implemented it in Matlab and embedded it in the ADN platform with system people.
- a Kalman filter based closed-loop control algorithm for fast control plane. Implemented it and verified the performance in Matlab.
- a filed patent on a joint optimization of traffic flow routing and network slicing.
- a filed patent on a network architecture based on the hybrid control planes.
2. New Traffic Blocking Theory for Data Services in Wireless Networks
07/2016-08/2017
Position: Project Leader, Future Network Theory Lab
Contributions:
Position: Project Leader, Future Network Theory Lab
Contributions:
- interactions with Huawei Wireless Product Dept. and China Telecom to clarify the challenges and requirements for wireless data network planning.
- processed the wireless traffic data collected by Huawei Wireless Product Dept. and verified the long-tail property of the data services using statistical fitting techniques.
- a mathematical framework to derive the probability density function of the queueing system under long-tail data traffic based on Lindley’s equation.
- a white paper on traffic blocking theory for data services at Huawei official platform.
- participated in the online software tool kit design for visualizing the results of the new traffic blocking theory.
3. Rate Control Algorithm Design in Data Center Networks
12/2015-04/2016
Position: Algorithm Designer, Future Network Theory Lab
Contributions:
Position: Algorithm Designer, Future Network Theory Lab
Contributions:
- a dynamic system modeling and a rate control scheme for deadline-sensitive flows and latency-sensitive flows in DCN.
- worked with a system team in Nanjing University to verify the performance of the proposed algorithms by comparing with traditional schemes.
- a published paper in INFOCOMM 2017.
4. Cooperative Projects with CalTech and CornellU
09/2015-08/2017
Position: Project Manager, Future Network Theory Lab
Contributions:
Position: Project Manager, Future Network Theory Lab
Contributions:
- worked together with PIs and Huawei Corporate Dept. from project proposal at the initial stage, monitoring at the middle stage, to project assessment at the final stage.
- a closed-loop network control algorithm with information sharing constraints in the collaboration with CalTech.
- an optimization-based rate control scheme for deadline-sensitive flows in DCN in the collaboration with CornellU.