The rapid growth of spectral resources’ demands, as well as the increasing Quality of Service (QoS) requirements of wireless users have led to the necessity for new resource allocation schemes which will take into account the differentiated QoS needs of each wireless user. Towards this direction, the researchers have introduced the concept of effective capacity, which is defined as the maximum rate that the channel can support in order to guarantee a specified QoS requirement. This concept has been considered as a “bridge” among the physical layer characteristics and the upper-layer metrics of QoS. During the last years, it has been widely employed for resource allocation problems in various wireless networks leading to efficient mechanisms. This chapter focuses on the employment of the effective capacity theory in Cognitive Radio (CR) systems, presenting an extensive survey on QoS-driven resource allocation schemes proposed in the literature. Some useful conclusions are presented and future research directions on this subject are highlighted and discussed.
Top1. Introduction
The emerging services of modern wireless networks raise an increasing demand for spectral bandwidth resulting in the congestion of this scarce resource. At the same time, the limited spectrum is used inefficiently most of the times leading to low utilization. To overcome these problems, the concept of CR technology has been introduced. The research community defines CR as a radio platform that can rapidly reconfigure its operating parameters through a process of cognition, based on changing requirements and conditions. It is considered that in a CR system, the licensed wireless spectrum can be opportunistically accessed by unlicensed users (Secondary Users, SUs) in an intelligent and flexible way, without causing harmful interference to the licensed users (Primary Users, PUs). The spectrum sharing techniques can be divided in the spectrum interweave, spectrum underlay and spectrum overlay (Goldsmith, Jafar, Maric & Srinisava, 2009) approaches. In the first approach, the SUs are allowed to transmit their data only in the absence of PUs’ transmission, whereas in the spectrum underlay approach, the SUs are allowed to transmit at the same time with the PUs guaranteeing specific interference constraints. Finally, in the spectrum overlay approach the SUs are allowed to transmit simultaneously with the PUs but they are obligated to help the PUs and relay their messages so as to offset the caused interference. Each of these approaches is characterized by its own challenges and benefits.
Furthermore, the fact that CRs can sense the environment and adapt their operating characteristics correspondingly, leads to their potential use in many applications and scenarios. One of the possible approaches for exploiting CR networking is the case of the Opportunistic Networks (ONs) (Stavroulaki et al, 2011) which constitute dynamically created networks and operate as extensions of the existing telecommunication systems. Given the dynamic environment of ONs, cognitive mechanisms should be proposed so as to allocate efficiently the network resources. Various resource management algorithms can be found in (Georgakopoulos et al, 2012; Karvounas et al, 2012) for a more detailed analysis based on the specific networks.
At the same time, apart from the number of the wireless users and the consumed spectral resources, the QoS requirements of the various services increase dramatically. Moreover, different services (such as video, VoIP, multimedia etc.) demand different QoS constraints leading the wireless users to tolerate different levels of delay for each service. Especially for real time applications, it is important to take into consideration the impact of the QoS provisioning metric in the system’s performance analysis. Therefore, there is an obvious need for new resource allocation schemes that will also take into account the quality requirements of the users. The combination of these schemes with the concept of CR will lead to much more efficient use of spectrum guaranteeing also the QoS satisfaction of each user. Typically, in a CR system, the SUs are either not allowed to transmit at all when a PU is present or they have to bound their transmit power so as not to harm the PU’s communication, depending on the spectrum sharing approach. This constitutes an additional barrier for the SUs in order to satisfy their QoS requirements, compared with the PUs whose transmission is mainly limited due to the propagation phenomena and intra-system interference constraints. Thus, the QoS provisioning is more challenging in CR systems and particularly for the SUs, since they also have to take into account the activity of the PU and the imposed interference constraints. Consequently, cognitive users have to allocate their resources properly in order to be able to achieve specific QoS requirements.