Multi-source multicast service chaining model guaranteeing reliability of network services

Abstract

Service chaining provides network services to users by processing packets with a series of virtualized network functions (VNFs). Multicast communication can suppress the network load compared to unicast communication when the same service is provided to multiple users simultaneously. Network providers need to ensure that services are provided to users under reliability requirements. Existing service chaining models can cause failure-intolerant or cost-inefficient resource allocation for multicast services. This paper proposes a multi-source multicast service chaining model that guarantees the reliability of network services with flexible routing and VNF placement. The proposed model determines the routing and VNF placement for each service chain request with redundancy so as to minimize the total resource utilization cost while satisfying the reliability requirement. In order to obtain cost-efficient feasible solutions, we introduce two heuristic algorithms. One is an algorithm that iteratively solves an integer linear programming problem until the reliability requirements are satisfied for all requests. The other is an algorithm that determines the VNF placement by using a metric based on the betweenness centrality and server failure rate. Numerical results show that the proposed model allocates the network and computation resources for services with lower total cost compared to the existing model. We also show illustrative examples of the resource allocation to compare the characteristics of two heuristic algorithms.