Summary and Future Recommendation on Cognitive Radio Network

Few friends have shown interest by requesting for more information about Performance Modelling & Analysis of Secondary Users in Cognitive Radio Network that I posted, so I thought I should share more about it.   

                                 Summary & Conclusion of the Project

    The project looked into a number of discrete event simulations to analyse the performance of SU users under different conditions. The traffic patterns and behaviour of SU and PUs are characterized by using PRP M/M/1, M/GE/1 and GE/GE/1 queuing network models, the comparison of simulation results of these models show that performance of SU user in CR network is affected by different traffic patterns in terms of inter-arrival and service times of both PU and SU calls. PRP M/GE/1 is modelled to study the impact of multiple transmission channels on the performance of CR users and results show that the PU performance degrade with increased number of channels. In the course of attempting to improve SU performance, effects of three different channel selection algorithms (shortest SU queue, random selection and next SU queue) are studied and results reveal that the shortest queue algorithm gives the best SU performance. Furthermore, two different spectrum hand off strategies (always stay and always change) are studied and results show that the

performance of SU in a multi channelled system is improved when they decide to change channel rather than stay in the SU queue of its current channel when pre-empted. For both real time and non-real time SU traffic, the performance measures studied are mean delay, throughput, mean number of SU in queue, forced termination probability, and call loss probability.

       Finally a comparative analysis between model proposed in this dissertation and that of an existing work is carried out and results show that the model proposed in this project gives better SU performance in terms of Mean system dwelling time and gives the same performance in terms of SU call blocking probability.

                                             Future Work

CR technology is working towards efficient use and management of radio spectrum in order to overcome the problem of spectrum scarcity. In this dissertation, the SUs are assumed to be aware of the state of the channels through some spectrum sensing algorithm, this work can be extended by incorporating and actually implementing spectrum sensing algorithms and also it can be used as a block to investigate performance of SU where they are allowed to share the channel with the PUs as long as they don’t go beyond some tolerance. This should reduce the frequency of pre-emptions by PUs and invariably improve SUs’ performance.