A Study Of WEB Service Composition Algorithm On Graph Model

At present,most web services composition algorithm focuse on finding the optimal QoS service composition,but a single service composition solution can not meet the user’s preferences and single choice is easy to bring performance bottlenecks to the system.At the same time,the dynamic change of the network environment leads to service failure,which makes the availability of single service composition scheme greatly reduced.Therefore,it is more effective and available to provide QoS-based Topk service composition for users.However,the existing QoS-based Topk service composition algorithm can not guarantee the accuracy and time performance of the composition results.Therefore,based on the service dependency graph,this paper proposes the QWSC-K algorithm to solve the Topk service composition problem.This algorithm uses the service dependency graph to construct the relationship between services,and then,different optimization strategies are adapted to reduce running time:Firstly,the front and backward level filtering are used to remove service which is irrelevant to the parameters of request service.Then,composition result is directly represented by combination path sequence to avoid the time-consuming combination backtracking process.Finally,in the process of searching traversal,the number and combined cost of combined sequences can be reduced by using the method of dynamic programming.At the same time,in order to ensure the accuracy of service composition results,the priority queue is used to save combined path sequences in the process of traversing and avoid the emergence of local optimal situation.At the same time,most of the existing service composition methods are adapted to obtain the QoS-optimal service compositions or approximate QoS-optimal service combinations in the static environment.However,the dynamic changing service environment on the Internet makes the original service composition unused due to any invalid service.At present,although there are some researchs on service composition algorithm for dynamic environment,most of them focus on the dynamic update of the optimal-QoS service composition and lack the specialized research on Topk service composition.Therefore,this paper base on the QWSC-K to propose the Topk service composition algorithm in dynamic environment DQWSC-K.This algorithm mainly uses the dynamic events of the distribution/subscription network to monitor and obtain the dynamic service and use dynamic services to update the original service dependency graphs.Then,on the updated service dependency graph,the incremental adaptive algorithm only updates the state of affected services without searching the entire services space.Compared with the existing methods,the DQWSC-K algorithm cache the middle traversal results and dynamically update the intermediate results to avoid inefficient recalculation,which can ensure the quality of service composition under the higher algorithm performance.Finally,in the static environment,the comparison experiment with the existing algorithms proved that the QWSC-K algorithm has advantages in time performance and accuracy of service composition results.In dynamic environment,through the comparison with the recalculation method,it is shown that the DQWSC-K algorithm has good performance advantage.