Management of cognitive radio ad hoc networks using a congestion-based metric Paul Potier * † and Lijun Qian Electrical and Computer Engineering Department, Prairie View A&M University, Texas A&M University System, Prairie View, TX 77446, USA SUMMARY In this paper, we propose a new network management protocol to address the unique challenges of managing cognitive radio ad hoc networks that have distributed, multihop architectures with dynamic spectrum availabil- ity. We focus on performance management for these networks, and address the problem of network congestion for secondary users, because of its significant impact on data throughput. Specifically, we define a performance metric, the average congestion level of the network, and derive it analytically as a function of the primary users’ activities and the secondary users’ strategy. For practical implementation, we further propose a cluster-based management architecture that utilizes a designated central manager and cluster heads that function as distributed managers. The cluster heads collect information from multiple layers of the protocol stack using new MIB (management information base) variables to capture the characteristics of cognitive radio ad hoc networks, such as the location-dependent spectrum availability. The objective of the management action is to utilize a network- level view of the congestion situation in the network by directing the secondary users to select the highest- quality links available and avoid congested clusters. This hierarchical networkmanagement design allows us to take advantage of its scalability to achieve near-real-time management. Numerical results demonstrate the effectiveness of the proposed scheme. Copyright © 2013 John Wiley & Sons, Ltd. Received 16 June 2012; Revised 7 June 2013; Accepted 19 June 2013 1. INTRODUCTION With the increased use of radio-frequency devices competing for scarce spectrum resources, the US Federal Communications Commission (FCC) has indicated in a 2002 report that portions of the spectrum are significantly underutilized [1]. The spectrum shortage and the inefficient usage of spectrum has encouraged the development of cognitive radio (CR), which is a context-aware intelligent radio capable of autonomous reconfiguration by learning and adapting to the spectrum environment [2,3]. The emergence of bandwidth demand for a host of wireless devices has motivated new spectrum allocation policies, which allow unlicensed secondary users (SU) to access the radio spectrum when it is not occupied by licensed primary users (PU). The intent is that this cognitive radio approach will improve spectrum utilization in wireless communications systems. Since the unlicensed or lower-priority SUs employing CR must limit any interference to the PUs, the SUs must only transmit in the spectrum holes left available by the PUs, where a spectrum hole is a frequency band assigned to a primary user that is not being used at a particular time and geographic location [3]. CR networks may be composed of single-hop networks, such as the fixed point to multipoint type as described by the IEEE 802.22 specification. They may also include multihop links, without a central base station, and these networks are called cognitive radio ad hoc networks (CRAHNs).They are char- acterized by dynamically changing network topology and spectrum availability. The major functions *Correspondence to: Paul Potier, Electrical and Computer Engineering Department, Prairie View A&M University, Texas A&M University System, Prairie View, TX 77446, USA. † E-mail: paulo99@swbell.net INTERNATIONAL JOURNAL OF NETWORK MANAGEMENT Int. J. Network Mgmt (2013) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/nem.1835 Copyright © 2013 John Wiley & Sons, Ltd.