Optimal allocation of transmit power and bits per channel use in competitive cognitive radio networks

Abstract

Cognitive radios (CRs) are considered as a possible enabling solution for Dynamic Spectrum Access (DSA) systems. In a DSA system, a cognitive radio adapts to the environment by sensing the spectrum and takes quick decision on appropriate trans-mission parameters to achieve certain performance goals. A cognitive radio network (CRN) is de ned as a network of cognitive radios/secondary users (SUs). In a CRN, the resource allocation method is responsible for avoiding harmful interference at the primary users (PUs) while optimally utilizing the available resources. Resource alloca-tion problem is usually based upon a system model. A competitive CRN corresponds to a system model where multiple SUs share a single channel and multiple channels are simultaneously used by a single SU to satisfy their bits/channel use requirements.

In this thesis, a competitive CRN is assumed and for such an environment a resource allocation optimization framework is proposed to determine the optimal transmit power and bits/channel use distribution for SUs with two objective functions - minimization of the total transmit power and maximization of the total bits/channel use, and set of constraints such as interference temperature threshold, power budget, quality of service (QoS) of SUs. An upper bound on probability of bit error and lower bound on minimum bits per channel use requirement are considered as QoS of the competing SUs. The users power budget is considered across channels to exploit better channel conditions and hence to improve bits/channel use capability of the resource allocation problem. An interference threshold constraint is considered in order to protect PU's transmission. Firstly, the proposed optimization framework is solved in a centralized manner, which shows that more transmit power is required

xix

xx

in a channel with higher noise power and bits/channel use increases with increasing signal to interference plus noise power ratio (SINR). Moreover, the simulation results also show that the framework is more capable of supporting high bits per channel use requirement than other existing frameworks. Finally, a user based distributed approach is developed to solve the proposed framework using \Game Theory." It is seen that user based distributed solution also follows centralized solution.