Research On P2P Task Scheduling Scheme Based On Improved ICSA

With peer-to-peer computing model Rising, the substantial increase in network bandwidth and Internet computing power increasing rapidly, how to make full use of these network resources, to construct a large-scale, highly scalable, highly reliable, high-performance distributed computing system in a dynamic peer-to-peer network environment, which research is a hot one in recent years. Task scheduling is a key technology in P2P computing, it is a direct impact on the computing performance of the entire system. Task scheduling problem is that assigning a group of parallel processing tasks to nodes in accordance with the timing regulations by certain scheduling strategy, with a view to obtaining a better system performance. As the issue can not obtain optimal solution in polynomial time, it has been recognized as an NP-complete problem.For the NP-complete problems, in recent years the rising immune algorithm (IA, Immune Algorithm) is a better solution, that is, in a short period of time to find a better solution. Many experts on distributed systems began concerning immune algorithm research. The algorithm have unique advantages in solving complex issues, such as large space, nonlinear, global search excellence and so on, but the traditional methods don't, so the IA in combinatorial optimization and task scheduling has fairly good applicationes, the IA theoretical researches also develop rapidly.On the basis of Clonal selection theory, we introduce the optimization algorithm opt-aiNET, on this basis; improved cloning immune algorithm is given. Using the several different properties testing function proposed by De Jong, its convergence is studies by experiment, and the test data is compared with the data of other methods in the literature, experiment result verifies that the ICSA improved algorithm has better convergence, stability and strong adaptability. Using orthogonal test and variance analysis techniques, reveals the control parameters' impact on both local convergence performance and global convergence performance, some principles of the parameters chosen is proposed.For some factors of task allocation and scheduling in P2P system, such as nodes' on-line time, nodes' performence, network topology and communication mechanisms and so on, both node location on improved Gnutella message mechanism and idle time statistics mechanism on immune learning and memory are given, to access nodes' activity state and performance parameters, by fitting parameters mechanisms to access better nodes. Because nodes join and leave randomly in P2P system, the choice strategy of acting nodes is given. The coupling task data structure based on DAG is introduced, taking artificial immune system as research background, improved immune clonal selection algorithm (Immune Clonal Selection Algorithm, ICSA) is applied into the distribution of both tasks and nodes. Appropriate antigens, antibodies and chromosome expression are given, "antibody choice", "antibody clone", "antibody reorganization" and "antibody variation" operators are designed. Theoretical analysis and simulation experiments show that the proposed method is better than both the traditional method and genetic algorithm in solving task scheduling problem, it has a immune memory characteristic.