| Summary: | With the rapid deployment of data rate hungry wireless services, the demand for the radio spectrum is increasing day-by-day. On the other hand, fixed spectrum access (FSA) policy that allocates/assigns a certain portion of radio spectrum to a certain group of users has traditionally been adopted. Recent studies on spectrum usage have revealed that a large portion of the allocated spectrum is underutilized. In order to improve the overall spectrum utilization, recently dynamic spectrum access (DSA) policy has received a great deal of attention. Cognitive radio (CR) is a key enabling technology in order to facilitate DSA. CR can sense the radio spectrum and
based on sensing outcome, it can adjust various transmission and operating parameters including bandwidth and power. Although CR technology in
conjunction with DSA policy can improve the overall spectrum utilization, there are a number of challenges in designing CR based networks. Since
the availability of radio spectrum for such networks is dynamic in nature, it is a quite challenging task to meet the quality of communications for the users in CR network (CRN). Another design challenge is spectrum sensing
as an imperfect spectrum sensing can lead to a disturbance/collision to the original users (referred to as primary users) of the spectrum currently using the spectrum. In order to address some of the above mentioned design challenges, in
this thesis we make two major contributions, as follows. First, we develop resource allocation mechanisms that allocate available transmission rate of a particular CR user among its different classes of services using a cross-layer design approach that jointly considers the time varying nature of communication channels, availability of spectrum, and data link layer quality requirements of different classes of services. In order to study the effect of imperfect sensing on data link layer's packet level performances as well as on collision probability, in the second part of this thesis, we develop a queuing analytic model that incorporates imperfect sensing. This analytic model is also useful for a call admission control decision in CRN when there is a certain sensing error as well as certain quality of service requirements for both primary and CR users. Using our developed model, we also compare performance of a random transmission protocol with that of the traditional deterministic transmission protocol.
|
|---|
