Distributed Packets Scheduling Technique for Cognitive Radio Internet of Things Based on Discrete Permutation Particle Swarm Optimization

Abstract

Integrating Cognitive Radio Networks with Internet of Things, which is called Cognitive Radio Internet of Things (CRIoT), is considered to be a breakthrough for the IoT spectrum scarcity problem. Some IoT applications can be controlled in a centralized manner while others needs a distributed manner. This depends on the nature and the architecture of the deployed IoT application. This paper proposes a distributed scheduling protocol to be applied on IoT for traffic control. Three distributed access strategies are examined; classical round robin and two proposed ones named vertex cover and enhanced round robin. They are used along with the centralized scheduling protocol (SDP-PSO) proposed by the authors in [1], to produce a distributed scheduling protocol. Dealing with this scheduling as an optimization problem; an objective fitness function is formulated, so that the fog node responsible to schedule at a time interval will formulate an allocation scheme for packets at itself. Using fairness index, average queuing delay and percentage of dropped packets as performance metrics, the performance evaluations of the three access strategies show that classical round robin performs better in case of using small number of fog node. However vertex cover performs better only in all other cases. Enhanced round robin performance is in the middle between the two other protocols.