P2P TASK SCHEDULING SCHEME BASED ON IMPROVED ICSA

摘要

In order to solve the problem of coupling task schedule in distributed P2P environment, the research on the scheme of coupling task scheduling is implemented by adopting the improved immune clonal selection algorithm. We present the mathematical model of coupling task schedule firstly, and the immune clonal selection algorithm is applied into the solution of this model. Antigen denotes the set of coupling task to be scheduled, antibody used to denote the outcome of effectively task scheduling. On which basis, an effective data structure of both antigen based on DAG and antibody that employs chromosome of three layer dynamic encoding is designed, and the minimum Makespan algorithm for P2P task scheduling are presented. Information-theoretic entropy used to maintain the variety of antibody to avoid “prematurely”, we concentrate on researching the immune clonal operator used to produce highly efficient task scheduling result. Experimental results show the algorithm that proposed in the paper outperforms the heuristic algorithms in scheduling mostly and performs pretty well in scheduling similar tasks and nodes frequently.