Skip to the content of the web site.

Cooperation in Cognitive Radio Networks

The radio spectrum and its use are strictly managed by governments in most countries, and spectrum allocation is a legacy command-and-control regulation enforced by regulatory bodies, such as the FCC in the United States and Ofcom in the United Kingdom. Most of the existing wireless networks and devices follow fixed spectrum access (FSA) polices to use the radio spectrum, which means that radio spectral bands are licensed to dedicated users and services, such as TV, 3G networks, and vehicular ad hoc networks. Licensed users are referred to as the primary users (PUs), and a network consisting of PUs is referred to as a primary network. In this context, only the PUs have the right to use the assigned spectrum, and others are not allowed to use, even when the licensed spectral bands are idle. Although interference among different networks and devices can be efficiently coordinated by using FSA, this policy causes significant spectral underutilization. The introduction of new wireless applications and services along with emerging diverse wireless networking architectures, e.g., heterogeneous networks, has been hampered by this inefficient allocation of radio spectrum.

Dynamic spectrum access (DSA) schemes which allow the unlicensed or secondary users (SUs) to share or reuse the licensed spectrum without interfering with the PUs have attracted considerable attention in both academia and industry. By using DSA, the SUs can (1) dynamically and opportunistically sense unused bands (spectrum holes) and transmit if spectrum holes are detected, (2) concurrently and transparently transmit in the licensed bands as long as the PUs are not interfered, or (3) cooperatively and trustfully negotiate with the PUs for transmission opportunities by providing tangible service. The first paradigm is referred to as spectrum sensing access (SSA), the second one is referred to as underlay coexisting access (UCA), and the third one is referred to as cooperative networking access (CNA), respectively. One key enabling technology of DSA is the cognitive radio (CR) equipped by the SUs, i.e., the network consists of CRs is a cognitive radio network (CRN).

In SSA, transmission durations of the SUs are highly random due to the randomness in the acquisition of spectrum holes, and transmission terminations when active PUs are sensed degrade performance, i.e., the quality of service (QoS) cannot be guaranteed. Moreover, the SUs should exert high energy and overhead to attain accurate spectrum sensing. In UCA, the interference temperature model is imposed on the SUs' power to avoid interference to the PUs, which leads to the SUs may only achieve short-range communications. In addition, the cumulative interference from multiple SUs to one PU and the interference from the PUs to the SUs deteriorate transmission performance of the PUs and SUs, respectively.

Unlike SSA and UCA paradigms wherein the PUs are unaware of the existence and needs of the SUs, the SUs are able to establish cooperation with the PUs via being relaying nodes for the PUs or by monetary leasing the licensed spectrum from the PUs in CNA. Specifically, in their roles as relays, the SUs can obtain transmission opportunities through providing multi-hop services, or providing additional propagation paths for the PUs; in their role as lessees, the SUs can exclusively use the licensed spectrum in short-term by paying the PUs. This shows transmission opportunities of the SUs are dedicated rather than random and QoS-guaranteed communications can be achieved by using CNA. CNA is also termed cooperative cognitive radio networking (CCRN).

Spectral efficiency and transmission performance can be improved by introducing user cooperation into CCRN. However, several challenging issues should be considered.

To establish a collaborative relation between different users, there must be motivation for partners to cooperate. Therefore, stimulating cooperation motivation is a prerequisite to establishing cooperation. Meanwhile, the metrics to evaluate cooperation performance should be addressed. We design two simple cooperation schemes that can stimulate both active and inactive PUs, and the SUs to participate in cooperation and achieve mutual benefits in CCRN. To generalize the evaluation metrics in terms of throughput optimization, a weighted sum throughput optimization problem is formulated.

To fulfill ideal cooperative communications, advanced physical-layer signal processing is indispensable. Synchronization between cooperators, orthogonal transmission and relaying must be carefully designed in a complicated time-varying transmission environment. We develop a simple and efficient orthogonal signaling scheme based on quadrature amplitude modulation (QAM) to attain orthogonal transmission between cooperators.

To achieve satisfied cooperation performance, finding the most efficient cooperator is a challenging issue. The MAC-layer protocol must coordinate the cooperator selection and management simply and efficiently. High communication overheads or collisions may degrade the cooperation performance. In addition, interference during the multi-user coordination should be avoided. To tackle this issue, a novel orthogonal multi-user coordination scheme is proposed.

To avoid co-channel interference when spectrum leasing is exploited, one spectral band is reused in different areas, i.e., like the spectrum reuse in cellular network. Therefore, spectrum reuse in CCRN is limited spatially. A promising method to further efficiently use temporarily unused spectrum is to cooperatively reuse the same spectrum when multiple SUs are within interfering range. We propose a cooperative leasing scheme for neighbouring SUs leasing one spectrum band together from an inactive PU and exploiting cooperative communications with each other. By doing so, co-channel interference among the SUs is avoided, and the transmission performance of secondary links and spectral efficiency are further improved.

To deploy user cooperation with high spectral efficiency in CCRN, the foremost issue is to exploit the available degrees of freedom in the wireless network, e.g., time, space, coding, modulation, etc, and how to efficiently manage and use these degrees of freedom is critical to the CCRN design and implementation. Another issue that should be taken into consideration is to stimulate cooperation motivation of the PUs and the SUs in CCRN, which means the user cooperation framework, should be based on the principle of achieving mutual benefit to all cooperators.

As shown in Fig. 1, we discuss two types of user cooperation models in CCRN. The first one is the cooperative communications between the SUs and active PUs, where the SU relays the PU's traffic and obtains a transmission opportunity as a reward. The other model is the cooperative communications among the SUs in which the SUs lease a band from an inactive PU together and establish cooperative communications with each other. We call this cooperative leasing in CCRN.

It is known that wireless communications are based on the polarization-sensitive electromagnetical (EM) medium, and polarization can be used as an independent domain. Exploring polarization is an exciting research direction and may bring revolutionary subversion fundamentally for wireless communications. It is expected that the design criterion and main performance metric of wireless communications will shift from time/frequency/space domains to time/frequency/space/polarization domains. Note that the time, frequency, and space domains have been exploited for CCRN. This shows that exploiting polarization for CCRN may be a logical next step. Unlike MIMO systems which require at least half wavelength in antenna spacings at terminal devices, there is no spacing requirement for placing the orthogonally dual-polarized antenna (ODPA), i.e., the schemes leveraging DoF provided by polarization are space- and cost-efficient. Furthermore, polarization signal processing is an additional method besides time, frequency, and space based methods. This shows that the use of polarization is a promising framework for CCRN.

In the considered networking scenario, taking the uplink communications as an instance, a PT communicates with the PR, and the STs need to communicate with the SRs, as shown in Fig. 2. If the primary network (PN) is a cellular network, then the primary transmitters (PTs) are battery-powered mobile terminals, e.g., cell phones, and the primary receiver (PR) is a base station powered by alternating current (AC). The PR allocates network resources to the PTs, e.g., frequency bands, time slots, spreading codes, etc, so that different PTs can communicate with the PR without interfering with each other. In the secondary network (SN), the secondary transmitters (STs) are battery-powered mobile terminals, and the secondary receivers (SRs) are access points (APs) powered by AC. Furthermore, the PTs and PR still use legacy uni-polarized antennas and are not required to change their hardware to support polarization capability, and the STs and SRs equip ODPAs. Furthermore, we consider both the PUs and the SUs cannot equip spatially distributed multiple antennas due to size limitation.

In Cooperative Cognitive Radio Networks (CCRN), the PU enhances its performance through cooperation with SUs. In return, the cooperating SUs can gain opportunities for its own transmission. However, almost all the related works assume that SUs are trustworthy and well-behaved, which may not be always true in reality. There may exist some dishonest users, even malicious ones in the system, corrupting or disrupting the normal operation of CCRN. Consequently, the performance can therefore be compromised. Thus, security also needs to be considered for this emerging cooperative networking.

Moreover, cooperation among multiple SUs has been introduced into the spectrum sensing system to deal with the factors of fading and shadowing. In cooperative sensing systems, the malicious SUs may send false information to mislead the spectrum sensing results, causing collision or inefficient spectrum usage, which is referred to as false sensing report attack; or the malicious SU transmits signals whose characteristics emulate those of PUs' signals to disturb the spectrum sensing, which is referred to as the primary user emulation (PUE) attack.

In this research, we aim to develop cooperative framework for CCRN and attack-proof cooperative sensing system, taking the security into considerations.

Current Researchers:

Bin Cao, Yujie Tang, Ning Zhang.