Real-Time Delivery of 4G Services With Cross-Layered and Power-Optimized Cognitive Radio Architecture

In this paper, we develop a cognitive radio (CR)-based architecture for fourth generation (4G) users that provides real-time delivery of service to its applications through the intelligent assignment of the CR spectrum. CRs work in an unlicensed network setting, so it is important that cognitive users utilize the spectrum opportunistically, i.e., without interrupting the licensed user's activity. To provide the opportunistic access for a cognitive 4G user, we propose an architecture that delivers the desired solution by following a novel five-step process. First, the cognitive (or secondary) user relays its throughput requirements from the application layer down to the media access layer by way of cross-layering. Second, the media access layer at its end senses the licensed spectrum and, based on the sensed parameters, characterizes the licensed (or primary) user's amount of spectrum occupancy into three different occupancy level types. Third, the licensed user's spectrum occupancy, categorized into three levels, is statistically represented by setting up a Markov model that evaluates the probability that the licensed channel could be found in a particular state at a given time instant for each level. Fourth, to provide opportunistic spectrum access to the secondary user, an optimization strategy is devised in the media access control layer that will bring face to face the throughput requirements obtained from the application layer and the steady-state probabilities for the licensed channel from the physical layer. Fifth, optimal performance metrics concerning the 4G user throughput and power are attained by solving the optimization problem. Results obtained are simulated against a variety of network parameters, and performance improvement through the proposed scheme is assessed by simulations conducted for a long-term evolution (LTE) system.