A novel dynamic channel allocation protocol based on data traffic characterization model in CR-IoT network

In multi-user cognitive-radio internet of things (CR-IoT) network, accurate estimations of data arrival are critical for secondary users to allocate channels. In the context of the data arrival model with long-term variations of rate, improving the accuracy of the performance evaluation of channel allocation protocols is an open issue. Thus, to evaluate the performance of various channel allocation protocols with predefined models of data arrival, a queuing analysis framework is developed using a probability allocation vector (PrA). The time-varying feature of data arrival is described by a Markov process including various data arrival states in the proposed framework. A dynamic probability allocation vector (DPrA) protocol capable of adjusting allocation strategy according to the arrival states by constructing the PrA is proposed. For comparative analysis, a maximum throughput allocation (MTA) protocol for conventional data arrival model is also evaluated under the proposed framework. Numerical results show that the DPrA protocol outperforms the MTA protocol in various performance metrics. Furthermore, the proposed modeling method for data traffic can provide convenience and effectiveness when designing channel allocation protocols in a CR-IoT network.