Weizhen Mao's Current Projects

Efficient RFID Protocol Design (Collaboration with Qun Li)

The "Internet of things" captures the vision that every object in the world can be tagged and connected through wireless or wired networks. Although RFID is one of most important enabling technologies in the Internet of things, a wide range of problems arise in RFID applications (e.g., inventory control, asset tracking, animal tracking, contactless payment). We have accomplished a number of results in RFID tag population estimation, RFID reading performance improvement, and secure RFID query and search protocol. In this project, we aim to further investigate a number of topics, including making RFID reading protocol more efficient and preserving privacy in RFID systems.

Representative papers
  • B. Sheng, Q. Li, and W. Mao, Efficient Continuous Scanning in RFID Systems, Proceedings of the 29th IEEE Conference on Computer Communications (INFOCOM), 1-9 (2010). pdf
  • H. Hao, B. Sheng, C. Tan, Q. Li, W. Mao, and S. Lu, Counting RFID Tages Efficiently and Ananymously, Proceedings of the 29th IEEE Conference on Computer Communications (INFOCOM), 1-9 (2010). pdf
  • B. Sheng, C. Tan, Q. Li, and W. Mao, Finding Popular Categories for RFID Tags, Proceedings of the Ninth ACM International Symposium on Mobile Ad Hoc Networking and Computing (MobiHoc), 159-168 (2008). pdf

Parallel Job Scheduling with Setup Times
Malleable parallel jobs can distribute their workload among any number of available processors in a parallel computer in order to decrease their execution time. In contrast, nonmalleable parallel jobs must use a fixed number of processors. The ideal execution time of a malleable parallel job with length p is p/k if it utilizes k processors. However, inherently serial code and parallel processing overhead (from process management, shared memory access and contention, communication, and/or synchronization) often prevent actual execution times from achieving this ideal. It is natural to consider this extra time as a type of setup time, a term commonly used by the scheduling community. We have derived an execution time function that takes both speedup (i.e., p/k) and setup time into account for the parallel execution of malleable jobs. We have tested the validity of our mathematical model through numerous experiments on large parallel systems. We are working on the study of various algorithms that can efficiently schedule malleable parallel jobs online under the proposed model.

Representative papers

  • J. Havill and W. Mao, Competitive Online Scheduling of Perfectly Malleable Jobs with Setup Times, European Journal of Operational Research, Vol. 187, 1126-1142 (2008). pdf
  • R. Dutton, W. Mao, J. Chen, and W. Watson, III, Parallel Job Scheduling with Overhead: A Benchmark Study, Proceedings of the IEEE International Conference on Networks, Architecture, and Storage (NAS), 326-333 (2008). pdf
  • R. Dutton and W. Mao, Online Scheduling of Malleable Parallel Jobs, Proceedings of the ISATED international Conference on Parallel and Distributed Computing and Systems (PDCS), 1-6 (2007). pdf

Last updated: September 8, 2011.