Jianxin Liao

A frequent pattern based framework for event detection in sensor network stream data

In this paper, we presented a frequent pattern based framework for event detection in stream data, it consists of frequent pattern discovery, frequent pattern selection and modeling three phases: In the first phase, a MNOE (Mining Non-Overlapping Episode) algorithm is proposed to find the non-overlapping frequent pattern in time series. In the frequent pattern selection phase, we proposed an EGMAMC (Episode Generated Memory Aggregation Markov Chain) model to help us selecting episodes which can describe stream data significantly. Then we defined feature flows to represent the instances of discovered frequent patterns and categorized the distribution of frequent pattern instances into three categories according to the spectrum of their feature flows. At last, we proposed a clustering algorithm EDPA (Event Detection by Pattern Aggregation) to aggregate strongly correlated frequent patterns together. We argue that strongly correlated frequent patterns form events and frequent patterns in different categories can be aggregated to form different kinds of events. Experiments on real-world sensor network datasets demonstrate that the proposed MNOE algorithm is more efficient than the existing non-overlapping episode mining algorithm and EDPA performs better when the input frequent patterns are maximal, significant and non-overlapping.

Two-Level proxy: the media streaming cache architecture for GPRS mobile network

Based on the technical background of media streaming over GPRS (General Packet Radio Service) mobile network, a novel media streaming cache architecture of two-level proxy is proposed for streaming service in GPRS network CCP (Central Cache Proxy) and UECP (User-End Cache Proxy) are applied to improve the performance of streaming service and to decrease the load of GPRS core network Key technical details correlated to this architecture are researched, such as: reliability design of proxy server, caching policy, and strategy of charging and billing Analyses and simulation experiments prove that this two-level caching proxy architecture can effectively optimize the performance of streaming service in GRPS network without upgrading the original GPRS network entities.

JCCM: Joint Cluster Communities on Attribute and Relationship Data in Social Networks

JCCM (Joint Clustering Coefficient Method) algorithm was proposed to identify communities which are cohesive on both attribute and relationship data in social networks. JCCM is a two-step algorithm: In the first step, it clusters tightly cohesive cliques as community cores and we proposed a heuristic method to identify community cores with a probabilistic guarantee to find out all community cores. In the second step, JCCM assigns the community cores and peripheral actors into different communities in a top-down manner resulting in a dendrogram and the final clustering is determined by our objective function, namely Joint Clustering Coefficient (JCC). To consider the power of actors in different roles in community identification, we defined two regimes of communities, namely union and autarchy. Experiments demonstrated that JCCM performs better than existing algorithms and confirmed that attribute and relationship data indeed contain complementary information which helps to identify communities.

Modeling and analysis for an enhanced three-tier dynamic location management in 3g

Location management is a key function in mobile communication systems to guarantee the mobile terminals to continuously receive services. In this paper, we study an enhanced movement-based location management scheme for 3G cellular networks where the home location registers, gateway location registers and visitor location registers form a three-tier hierarchical mobility database structure. Furthermore, we formulate analytically the cost model of location update and paging for the proposed scheme and compare the proposed scheme with the basic movement-based location management scheme. The outcomes show that the enhanced movement-based location management scheme outperforms the other scheme.