Mathematical model for predicting forwarding rule counter values in SDN

The paper presents the mathematical model for SDN operation that describes counter values as a network flow function in the rule dependency graph. The basic properties of this model are explored and it is demonstrated that this model can be used to create an algorithm for counter values prediction. SDN provides a mechanism for network statistics analysis through counters that measure the number of packets processed by each forwarding rule. One problem with this mechanism is that in order to get complete network statistics, SDN controller has to send a lot of counter requests thus imposing a large overhead on the control channel and network devices. Thus, in order to minimize network load due to such requests, we need an algorithm that can predict counter values for each forwarding rule based on the minimum number of chosen counters.