Events
Jie Liu
Microsoft
Microsoft Research Senior Researcher
Redmond, WA, USA
Dr. Jie Liu is a Senior Researcher at Microsoft Research (MSR) and the manager of the Networked Embedded Computing group, which conducts fundamental and applied research on sensing and energy-efficient computing. His research interests are on sensor networks, mobile computing, and cloud computing resource management. He has published more than 50 papers in these fields. He is an Area Editor of IEEE Tran. on Mobile Computing and has organized numerous sensor network and pervasive computing conferences. Dr. Liu received his PhD on Electrical Engineering and Computer Science from UC Berkeley on 2001.
Abstract
Sensing Founda tions for Environmental Science
Sensors, especially those connected directly or indirectly to the Internet, give us unprecedented capabilities to interact with the physical world. Environmental science is one of the early adopters of wireless sensing network technologies, and the one presents some of the distinct challenges and drives some of the most exciting technology breakthroughs.
Over the years, the Networked Embedded Computing group at Microsoft Research has conducted fundamental and applied research on networked sensing with the foci on sensor network node discovery, interference resilience, and data reliability; sensor data management and analysis; online data visualization and coordination; and data semantics abstraction and programming models. Many technologies are motivated by environmental monitoring and related applications. For example,
• Data Center Genome is a project that study power efficiency in data centers through monitoring, modeling, and controlling server and cooling power consumption, heat distribution, and workload placement. In the project, we developed RACNet (Reliable ACquisition Network) [1]for reliable data collection that scale to high network density (e.g. hundreds of nodes within a 1 communication hop). (c.f. http://research. microsoft.com/en-us/projects/dcgenome/)
• SenseWeb is a project with a goal of creating a community data coordination and visualization environment on the web. Loosely coupled scientific research group, as well as sensor hobbyists, can share their data over a common, geo-centric platform, and visitors can visualize the data over space and time directly on the web. (c.f. http://research.microsoft.com/en-us/projects/senseweb/default.aspx)
• Cypress is a data compression and management framework for massive data streams. Motivated by cloud service data centers that collect up to a Terra-byte of performance per day, we studied how to compress and achieve the data with efficient diskspace [4]; how to answer common queries directly using the compressed without reconstructing the original signals [2]; and how to approximate correlation query results efficiently to trim down search trees [3].
Currently, we are also researching on adding mobility to networked sensing systems. Mobility increases system coverage and flexibility, but also introduces energy, network formation, and location sensing challenges. Some of our current work includes
• WiFlock: a ultra-low duty cycle mobile node discovery protocol that can achieve 0.2% wakeup duty cycle with less than 4 minutes node discovery latency;
• LEAP: a ultra-low duty cycle GPS receiving technology that can greatly reduce the energy consumption of GPS-based location tagging;
In addition, we are also interested in energy scavenging technologies, mobile phone based sensing, and compressive sensing technologies. We believe these technologies can enrich typical static wireless sensor networks with opportunistic, participatory, or animal enabled sensing, and by doing so reduce the system deployment cost, yet increase the overall data collection throughput.
We look forward to collaborating with environmental scientists to advance the state of art in networked sensing and apply it to one of the most critical research fields with the goal of understanding and better preserving the world’s environment.
References:
[1] Chieh-Jan Liang, Jie Liu, Liqian Luo, Andreas Terzis, and Feng Zhao, “RACNet: A High-Fidelity Data Center Sensing Network,” in Proceedings of The 7th ACM Conference on Embedded Networked Sensor Systems (SenSys 2009), Association for Computing Machinery, Inc., November 2009
[2] Galen Reeves, Jie Liu, Suman Nath, and Feng Zhao, “Managing Massive Time Series Streams with Multi- Scale Compressed Trickles,” in VLDB ‘2009: Proceedings of 35th Conference on Very Large Data Bases , Very Large Data Bases Endowment Inc., August 2009
[3] Abdullah Mueen, Suman Nath, and Jie Liu, “Fast Approximate Correlation for Massive Time-series Data,” ACM SIGMOD 2010.
[4] Charles Loboz, Slawek Smyl, and Suman Nath, “DataGarage: Warehousing Massive Performance Data on Commodity Servers,” VLDB 2010.