Spectrum Aggregation in Cognitive Radio Access Networks from Power Control Perspective

Introduction

The vast proliferation of wireless devices and services for mobile communications, WiFi and TV broadcast serves as a characteristic example of how much the society depends on radio spectrum which is becoming increasingly scarce. Towards this direction, Federal Communications Commission (FCC) (Federal Communications Commission, 2012) has been investigating the exploitation of innovative radio design techniques to optimally manage the radio resource frequencies. In addition, FCC has recommended over the past years that spectral efficiency could be significantly improved by deploying wireless devices which can coexist with the incumbent licensed (primary) users. Thus, motivated by the increasing needs for efficient spectrum usage, 3GPP has defined three functionalities that increase the provided capacity to the users, namely, Spectrum Aggregation (or Carrier Aggregation), enhanced use of multi-antenna techniques (MIMO), and support for Relay Nodes (RN). These functionalities have been introduced in the Long Term Evolution Advanced (LTE-advanced) so as 3G devices would exploit unused component carriers, in order to increase bitrate and to achieve efficient heterogeneous network planning. The Spectrum Aggregation increases the users’ capacity by providing to them more spectrum carriers (licensed or unlicensed) through novel spectrum sensing and spectrum allocation techniques that comprise admission control, power control and user partitioning algorithms. Enhanced use of multi-antenna techniques is used to increase the overall bitrate through transmission of at least two different data streams on at least two different antennas—using the same resources in both frequency and time, separated only through the use of different reference signals—to be received by two or more antennas. RNs are small-sized base stations operating at low power levels; they are used for providing enhanced coverage and capacity at cell edges, or connection to remote areas without fiber infrastructure. RN is expected to be connected to the Evolved Node B (eNodeBs) via a radio interface, whilst the radio resources are shared among the User Equipments (UEs) which are served directly by either the Donor eNodeB or the RN. Two types of RNs have been identified, based on whether they use the same frequency as the Donor eNodeB or not. In the former type, RN could suffer from self-interference issues which could be surpassed with a time sharing scheme between transmitting and receiving, or by placing the transmitter and the receiver at different locations. The RNs enable the efficient network planning by providing enhanced coverage and capacity at cell edges; RNs can also be used to connect to remote areas without fiber connection. Out of the three functionalities Spectrum Aggregation has received the most attention as the most straightforward way to increase the capacity offered to the users is to increase the available bandwidth.

The Spectrum Aggregation scenario considers as prerequisite a set of functionalities for efficient usage of the available resources. A typical cognitive radio network comprises of nodes that attempt to transmit opportunistically in unused frequency bands that are not licensed to operate at; spectrum sensing realizes the identification of the potential spectrum holes that a CR user could exploit. The admission control mechanism aims at identifying whether a CR user can be served using the available network resources, without causing interference to existing incumbent ones. However, unlicensed users also compete with each other for resource allocation, thus a power control mechanism that will address the interference mitigation among opportunistic users is also required. The spectrum sensing and the admission control functionalities are topics well investigated in the literature, whereas the opportunistic usage of resources has recently started attracting the interest of the research community, especially in such scenarios.