and for 1+1 DetNet flows

Abstract

Deterministic Networking (DetNet) is attracting a lot of attention lately due to their ability to provide bounded latency and zero packet loss for time-sensitive applications. In this paper, we formulate the routing and scheduling problem for 1+1 protected DetNet flows based on Cycle Specified Queuing and Forwarding (CSQF). The solution to this problem selects two paths between the two endpoints of each service request and schedules packet transmission on these paths meanwhile maximizing the accepted traffic. We have modelled the problem using Integer Linear Programming (ILP). We also propose two heuristic approaches: greedy and Tabu-search (TS) that can perform 1+1 routing and scheduling for a large number of requests in a reasonable time. Eventually, the performance of the ILP approach and heuristics is evaluated by performing simulation experiments. The results highlight the scalability of the two heuristics as compared to the ILP and superior performance of TS over greedy. The trade-off of cycle time on traffic acceptance and end-to-end delay is also presented.