Cooperative communication for cognitive radio networks

Abstract

A Cooperative Cognitive Radio (CCR) network, which integrates a conventional cooperative system and cognitive radios (CRs) into a holistic system, is a promising paradigm for the next generation mobile communication systems. The spectral efficiency, power efficiency, bandwidth reduction and system complexity in CCR networks are the fundamental parameters of our system design and optimization. In this thesis, we focus our attention on opportunistically exploiting the underutilized spectrum band in the CR network with the aid of cooperative protocols. Furthermore, we invoke channel coding schemes in our CCR system in order to improve the overall system throughput. In our CCR system, the overlay based cooperation scheme of Primary Users (PUs) and Cognitive Users (CUs) is considered, which has the potential of leading to a transmission power reduction and transmission rate improvement for both the PU and the CU. More explicitly, our cooperative protocol allows a group of CUs to serve as Relay Nodes (RNs) for relaying the signal of the PUs’ transmitters to the PUs’ intended destinations. To elaborate further, both one way relaying and two-way relaying schemes are used in our proposed system, so that the bandwidth requirement of the PUs is reduced. Alternatively, the freed bandwidth may be leased to a group of CUs for their secondary communications.

Our numerical and simulation results show that the bandwidth reduction attained by the proposed two-way relaying based CR scheme may approach as much as 80% of the PU’s bandwidth. Moreover, an Adaptive Dynamic Network Coding (ADNC) scheme is also conceived for this overlay CCR system, which is designed for supporting communications between multiple PUs and a common Base Station (BS). More particularly, the near-instantaneously Adaptive Turbo Trellis Coded Modulation (ATTCM) is employed for appropriately adjusting both the modulation mode as well as the channel coding rate and the network coding rate, according to the near-instantaneous channel conditions. In order to facilitate the recovery of the source information at the BS, the CUs invoke the ADNC technique, which is assisted by our cooperative protocol operating by exchanging the CCR-based control information between the near-instantaneously ATTCM and network coding codec as well as between the CUs and the BS. Additionally, the network encoder may also be activated in its adaptive mode for supporting the CUs, depending on the Boolean value of the feedback flags generated based on the success/failure of the ATTCMdecoder and of the network decoder, which is evaluated and fed back by the BS. Quantitatively, it was found that the joint holistic design of our ATTCM-ADNC-CCR scheme is either capable of freeing up an approximately 40% of the PU’s bandwidth in comparison to its non-cooperative counterpart, or increasing the attainable throughput by as much as 2 bit/symbol.

Furthermore, a Pragmatic Distributed Algorithm (PDA) is proposed for supporting the efficient spectral access of multiple PUs and CUs in CCR networks. The novelty of our PDA is that the Pus negotiate with the CUs concerning the specific amount of relaying and transmission time, and the CU will decide either to accept or to decline this offer. These CUs relay the signal received from the PUs to the PUs’ receiver, but only when both the PUs’ and the CUs’ minimum rate requirements are satisfied. Moreover, we show that the cooperative spectral access based on our PDA reaches an equilibrium, when it is repeated for a sufficiently long duration. These benefits are achieved, because the PUs are motivated to cooperate by the incentive of achieving a higher PU rate, whilst defecting from cooperation can be discouraged with the aid of a limited-duration punishment. Therefore, our proposed PDA outperforms the benchmark, despite its significantly lower overhead and lower complexity. Finally, we present the joint design of coding, modulation, user-cooperation and CR techniques, which may lead to significant mutual benefits for both the PUs and the CUs.

More information

Identifiers

ORCID for Soon Xin Ng: orcid.org/0000-0002-0930-7194

Catalogue record

Export record