Decentralized Machine Learning Intelligence > Topics > Adaptive Services Management

Adaptive Service and Resource Management


In recent years, inspired by the emerging web services standards and peer-to-peer technology, a new federated service computing model has received increasing research interests. The resulting system is what we called the federated service providing system (FSP) which exploits the popular P2P paradigm to achieve scalable provision of domain-specific services in large networked environments. One fundamental problem that determines the eventual success of FSP systems is how to achieve adaptive management of distributed services and resources in such systems where each node enjoys high level of autonomy. Without relying on a centralized controller which presents a serious performance bottleneck and single point of failure that affect the overall scalability of the system, we seek to manage services and resources in a self-organized manner through completely localized control. Specifically, we need to (1) properly and accurately model the behaviour of a FSP system and its performance; (2) support efficient and fast discovery of desirable services and resources in the FSP system; (3) develop fully distributed protocols that govern peer’s local interactions in the system such that global performance of the FSP system can be continuously improved; and (4) develop self-organization algorithms for systematic improvement of global performance. Research efforts have been made over the past few years to try to achieve these objectives.



Related articles:
  • G. Chen, A.Sarrafzadeh, C. P. Low and L. Zhang, "A Self-Organization Mechanism Based on Cross-Entropy Method for P2P-Like Applications," ACM Trans. Auton. Adapt. Syst., vol. 5, no. 4, pp. 15:1--15:31, 2010.|PDF|Bibtxt|
  • G. Chen, C. P. Low and Z. H. Yang, "Extremal Search of Decision Policies for Scalable Distributed Applications", 2nd International Conference on Scalable Information Systems, pp.77:1-77:8, 2007
  • G. Chen, C. P. Low and Z. Yang, "Coordinated Services Provision in Peer-to-Peer Environments," IEEE Transactions on Parallel and Distributed Systems, vol. 19, no. 4, pp. 433-446, 2008.|PDF|Bibtxt|
  • G. Chen, C. P. Low and Z. H. Yang, "Enhancing Search Performance in Unstructured P2P Networks based on Users' Common Interest", IEEE Transactions on Parallel and Distributed Systems, vol. 19, no. 6, pp. 821 -836, 2008