Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 106 === With the rapid development of Internet, the host-to-host IP Internet architectures cannot satisfy the requirement of Internet. Therefore, the content-centric networking (CCN) emerges to replace the traditional IP networking. Because the pending interest table (PIT) has a significant effect on the performance of the CCN network, we focus on the research on performance evaluation of the PIT built on the cellular network. In this work, we study the four scenarios for PIT occupancy: new interest without retrial, new interest with retrial, new and handoff interest without retrial, new and handoff interest with retrial. In new interest without retrial or with retrial scenarios, we focus on the cases where there is a single cell or all nodes are stationary and thus there are only
one class of arrivals: new interest. In the scenarios without retrial, if all entries are
occupied, an arriving interest is cleared from the system immediately whereas in the scenarios with retrial, an arriving interest may enter into a retrial queue and retry to find an idle entry later. In new and handoff interest without retrial or with retrial scenarios, we focus on the cases where there are multiple cells and all nodes are mobile and thus there are two classes of arrivals: new interest and handoff interest. First, we derived the analytical models for the system considered. An iterative algorithm is developed to find the steady state probability distribution and the performance measures of interest. The system parameters include interest arrival rate, interest service rate, interest timeout time, interest aggregation probability, interest impatient rate, interest dwell rate, interest retrial rate, and interest retrial probability. The performance measures of interest are interest blocking probability, average number in service, average system delay,throughput, and incomplete rate. Second, for comparison, we study the performance with retrial and without retrial and the difference between the cases in a single cell or multiple cells. Third, it is shown that performance with retrial outperforms that without retrial in terms of the throughput and the incomplete rate at the expense of the average system delay and blocking probabilities. Last but not least, in most cases studied the analytical results are shown to be in good agreement with the simulation results.
|