Title:Flow Sampling: Network Monitoring in Large-Scale Software-Defined IoT Networks

Abstract:Software-defined Internet-of-Things networking (SDIoT) greatly simplifies the network monitoring in large-scale IoT networks by per-flow sampling, wherein the controller keeps track of all the active flows in the network and samples the IoT devices on each flow path to collect real-time flow statistics. There is a tradeoff between the controller's sampling preference and the balancing of loads among devices. On the one hand, the controller may prefer to sample some of the IoT devices on the flow path because they yield more accurate flow statistics. On the other hand, it is desirable to sample the devices uniformly so that their energy consumptions and lifespan are balanced. This paper formulates the flow sampling problem in large-scale SDIoT networks by means of a Markov decision process and devises policies that strike a good balance between these two goals. Three classes of policies are investigated: the optimal policy, the state-independent policies, and the index policies (including the Whittle index and a second-order index policies). The second-order index policy is the most desired policy among all: 1) in terms of performance, it is on an equal footing with the Whittle index policy, and outperforms the state-independent policies by much; 2) in terms of complexity, it is much simpler than the optimal policy, and is comparable to state-independent policies and the Whittle index policy; 3) in terms of realizability, it requires no prior information on the network dynamics, hence is much easier to implement in practice.