Routing Protocols for Cognitive Radio Networks: A Survey

Transcription

1 Routing Protocols for Cognitive Radio Networks: A Survey 153 Routing Protocols for Cognitive Radio Networks: A Survey S. M. Kamruzzaman, Dong Geun Jeong 1 Abstract: with the cognitive radio (CR) technology, a wireless system can exploit opportunistically the radio spectrum licensed to other systems. Thus, CR is regarded as a solution to the problems resulting from the limited available spectrum and the inefficiency in the spectrum usage. The multi-hop CR networks need some novel routing algorithms taking the open spectrum phenomenon into account. The main approach in designing the routing algorithm for the CR networks is the joint design of routing and spectrum management. Works on such issues have just started and are still in a rudimentary stage. In this paper, we survey comprehensively the existing research on the routing protocols for CR networks, especially with reference to CR ad hoc networks. We classify the routing protocols, discuss the essential features of the different protocols, and provide the future research directions. 1. Background Cognitive radio (CR) technology has been proposed as a promising solution to address the increasing congestion in the unlicensed band by using the vacant spectrum of licensed band opportunistically while avoiding disruptions to the legacy users, such as, TV broadcast stations, public safety broadcast stations, wireless microphones, etc. The CR user or secondary user () is allowed to use only locally unused spectrum opportunistically; so that it does not cause any interferences or collisions to the incumbent or primary users (PUs) [1]. When CR users detect the presence of PU on the operating band, they must switch to other spectrum bands. The CR network (CRN) refers to the wireless network using CR technology [2]. CRNs can be classified as the infrastructure-based CRNs (ICRNs) and the CR ad hoc networks (CRANs). ICRN has a central controller such as base station in cellular networks or an access point in wireless local area networks (WLANs). In contrast, there is no infrastructure backbone in CRAN. Thus, a CR user can communicate with other CR users through ad hoc connection on both licensed and unlicensed spectrum bands [3], [4]. 1 Authors are with the Department of Electronics Engineering, Hankuk University of Foreign Studies, Korea ( {smzaman, dgjeong}@hufs.ac.kr) This work was supported by the Hankuk University of Foreign Studies Research Fund.

2 154 Journal of Information Industrial Engineering, Vol. 16, August 2010 Recent years have seen significant increase in the use of multi-hop wireless networks that operate in unlicensed spectrum using standardized wireless technologies (e.g., IEEE [5] based ad hoc networks and mesh networks). Meanwhile, it is well known that multi-hop networks suffer from rapid degrading performance as the number of users increases in the network. The fundamental impediment to building large multi-hop networks has been the insufficient network capacity, due to the limited spectrum available for unlicensed use [6]. Most of the research on CRNs to date has focused on single-hop scenarios, tackling PHYsical (PHY) layer and/or Medium Access Control (MAC) layer issues, including the definition of effective spectrum sensing, spectrum decision and spectrum sharing techniques [7]. Only very recently, the research community has started realizing the potentials of multi-hop CRNs which can open up new and unexplored service possibilities enabling a wide range of pervasive communication applications. Indeed, the cognitive paradigm can be applied to different scenarios of multi-hop wireless networks including cognitive wireless mesh networks featuring a semi-static network infrastructure [8] and CRANs characterized by a completely self-configuring architecture [9]. To fully unleash the potentials of such networking paradigms, new challenges must be addressed and solved. In particular, effective routing protocols must be integrated into the work already carried out on the lower layers (PHY/MAC), while accounting for the unique properties of the cognitive environment. In CRNs with multi-hop communication requirements, the dynamic nature of the radio spectrum calls for the development of novel spectrum-aware routing algorithms. In fact, spectrum occupancy is location-dependent, and therefore in a multi-hop path available spectrum bands may be different at each relay node. Hence, in multi-hop cognitive radio networks controlling the interaction between the routing and the spectrum management functionalities is of fundamental importance. While cross-layer design principles have been extensively studied by the wireless networking research community in the recent past, the availability of cognitive and frequency agile devices motivates research on new algorithms and models to study cross-layer interactions that involve spectrum management-related functionalities. In the remainder of the paper, we focus on the issues related to the design and maintenance of routes in multi-hop CRNs. The purpose of this work is twofold: First, we aim at dissecting the most common approaches to routing in CRNs, clearly highlighting their design rationale, and their strengths/drawbacks. Then, by leveraging the literature in the field, we comment on possible future research directions.

3 Routing Protocols for Cognitive Radio Networks: A Survey 155 PU Band 2 C 2 PU Band 1 C 1 PU Band 4 C 4 PU Band 3 C 3 Fig. 1. Routing in multi-hop CRNs. 2. Routing in Multi-hop Cognitive Radio Networks Comparing to traditional multi-channel networks, routing for utilizing multiple channels in multi-hop cognitive radio networks face several new challenges. Firstly, unlike multiple channel networks, the set of channels available for each node is not static. A unique challenge is the collaboration between the route selection and the spectrum decision. Due to the dynamically changed and intermittent spectrum band, the spectrum information is required when selecting the route. In order to adapt this variation, routing in multi-hop CRNs must be spectrum aware. Secondly, the lacks of a fixed common control channel (CCC). Since a CR user has to vacate the spectrum band as soon as a PU begins to use the network, the implementation of a fixed CCC becomes infeasible for CRNs. Thirdly, the spectrum-adaptive route recovery [10] process. In addition to node mobility, link failure in multi-hop CRNs may happen when PU activities are detected. How to vacate the current spectrum band and to move to another available spectrum band quickly is still an unexplored problem. Fourthly, the evaluation metrics for routing with channel assignment are still open issues. This makes how to deal with channel switching is a debatable question. Finally, the route maintenance/reparation; the sudden appearance of a PU in a given location may render a given channel unusable in a given area, thus resulting in unpredictable route failures, which may require frequent path rerouting either in terms of nodes or used channels. In this scenario, effective signaling procedures are required to restore broken paths with minimal effect on the perceived quality.

4 156 Journal of Information Industrial Engineering, Vol. 16, August 2010 Cognitive Radio Routing Protocols Global Spectrum Information Local Spectrum Information Graph Based MILP-MINLP Optimization Interference and Power Based Delay Based Tree Based Probabilistic Approach Throughput Based Link Quality /Stability Based Fig. 2. Classification of cognitive radio routing protocols. The reference network model reported in Fig. 1 features secondary devices which share different spectrum bands with PUs. Several spectrum bands (1,..., M) may exist with different capacities C 1, C 2, C M, and the s may have different views of available spectrum bands due to inherent locality of the spectrum sensing process. Typically the PUs are assumed motionless while the s may vary their position before and during a transmission. In this scenario, the problem of routing in multi-hop CRNs targets the creation and the maintenance of wireless multi-hop paths among s by deciding both the relay nodes and the spectrum to be used on each link of the path. Such problem exhibits similarities with routing in multi-channel, multi-hop ad hoc networks and mesh networks, but with the additional challenge of having to deal with the simultaneous transmissions of the PUs. 3. Classification of Cognitive Radio Routing Protocols In the following sections of this paper, several routing protocols are commented keeping an eye on the aforementioned challenges. In Fig. 2, we broadly categorize the proposed protocols into two main classes depending on the assumptions taken on the issue of spectrum-awareness: Global Spectrum Information; Local Spectrum Information. In the former case, a spectrum occupancy map is available to the network nodes, or to a central control entity, which could be represented by the centrally-maintained spectrum data bases recently promoted by the FCC to indicate over time and space the channel availabilities [11] in the spectrum below 900 MHz and around 3 GHz. The considered architectural model is a static cognitive multi-hop network where the spectrum availability between any given node

5 Routing Protocols for Cognitive Radio Networks: A Survey 157 pair is known. The routing protocols building on this assumption leverage theoretical tools to design efficient routes, differentiating on the basis of which kind of theoretical tool is used to steer the route design. A first class encompasses all protocols based on a graph abstraction of the CRN. The second sub-class instead employs mathematical programming tools to model and design flows along the cognitive multi-hop network. On the other hand, routing schemes based on local spectrum information include all those protocols where information on spectrum availability is locally constructed at each through distributed protocols. Thus, the routing module is tightly coupled to the spectrum management functionalities. Indeed, besides the computation of the routing paths, the routing module should be able to acquire network state information, such as currently available frequencies for communication and other locally available data, and exchange them with the other network nodes. The network state is primarily a function of node mobility and traffic carried in the network, network state in multi-hop CRNs is also influenced by PU activity. 4. Global Spectrum Information Based Routing Protocols Before sending or receiving data, cognitive opportunistic devices will be required to access these databases to determine available channels. Under this scenario, the central availability of up-to-date information on spectrum occupancy completely decouples the spectrum assessment modules (sensing, sharing) form the routing decisions/policies which can be locally optimized. This section comments on those routing protocols which start off from the assumption of global information on the spectrum occupancy, further proposing analytical tools to optimize/steer the routing decisions Graph-based Routing Protocols The general approach for designing routes in wireless multi-hop networks consists of two phases: graph abstraction and route calculation. Graph abstraction phase refers to the generation of a logical graph representing the physical network topology. The outcome of this phase is the graph structure G = (N, V, f(v)), where N is the number of nodes, V is the number of edges, and f(v) the function which allows to assign a weight to each edge of the graph. Route calculation generally deals with defining/designing a path in the graph connecting source destination pairs.

6 158 Journal of Information Industrial Engineering, Vol. 16, August 2010 Classical approaches to route calculation widely used in wired/wireless network scenarios often resort to mathematical programming tools to model and design flows along multi-hop networks Layered-graphs Routing The very same two-phase approach to route design has been leveraged also for multi-hop CRNs. The authors of [12] propose a comprehensive framework to address channel assignment and routing jointly in semi-static multi-hop CRNs. In these works, the PU dynamics are assumed to be low enough such that the channel assignment and the routing among s can be statically designed. The authors further focus on the case where cognitive devices are equipped with a single half-duplex cognitive radio transceiver, which can be tuned to M available spectrum bands or channels. The proposed framework is based on the creation of a layered graph which features a number of layers equal to the number of available channels. Each secondary user device is represented in the layered graph with a node, A, and M additional subnodes, A 1, A 2,..., A M, one for each available channel. The proposed layered graph is a rather general framework which can be combined with different routing metrics. Further modifications to the layered graph can be introduced to account for specific requirements. The layered graph framework is indeed useful to jointly model channel assignment and routing in semistatic multi-hop CRNs, where the topology variability dynamics is low. On the down side, the proposed path-centric routing approach is fundamentally centralized requiring network-wide signalling support to generate the layeredgraph. Moreover, the proposed iterative algorithm is suboptimal being based on a greedy approach Colored-Graphs Routing A similar approach based on graph structures is proposed in [13], where a colored graph is used to represent the network topology. The colored graph G c = (N c,v c ), where N c is the vertex set (one vertex for each network device), and V c is the edge set. Two vertices in the colored graph may be connected by a number of edges up to M, where M is the number of channels (colors) available for transmission on the specific link. Referring to Fig. 3 (a), Fig. 3 (b) corresponds to the colored graph abstracting the physical network topology. This approach obviously shares the very same drawbacks as the previously commented one. Namely, the

7 Routing Protocols for Cognitive Radio Networks: A Survey 159 Ch1 2 Ch1 Ch2 Ch2 1 4 Ch1 Ch1 Ch2 3 Ch2 1 Ch1 Ch1 Ch2 Ch2 2 3 Ch1 Ch2 Ch2 Ch1 4 (a) Network topology (b) Colored graph Fig. 3. Colored-graph creation. proposed protocol is centralized and heuristic, meaning that it may lead to suboptimal routing Optimization Routing Protocols In [14], [15], Hou et al. focus on the problem of designing efficient spectrum sharing techniques for multi-hop CRNs. They introduce a Mixed Integer Non-Linear Programming (MINLP) formulation whose objective is to maximize the spectrum reuse factor throughout the network, or equivalently, to minimize the overall bandwidth usage throughout the network. The proposed formulation captures all major aspects of multi-hop wireless networking. Mathematical programming is leveraged also in [16], where a Mixed Integer Linear Programming (MILP) formulation is derived for the problem of achieving throughput optimal routing and scheduling for secondary transmissions. The objective function aims at maximizing the achievable rate of source destination pairs, under the very same interference, capacity and routing constraints as defined above. The authors directly use the formulation to design route/channel assignment patterns for small-to-medium size network scenarios by resorting to commercial solvers. 5. Local Spectrum Information based Routing Protocols This section overviews those routing protocols where the retrieval of information on spectrum occupancy is performed in a distributed way, and, similarly to classical ad hoc networks, distributed approaches are introduced to make local radio resource management decisions on partial information about the network state. In multi-hop CRNs, such functionality

8 160 Journal of Information Industrial Engineering, Vol. 16, August 2010 is crucial since the local spectrum conditions acquired via radio sensing can be highly variable in time and space. The presented protocols are categorized according to the specific metric used to assess route quality Interference- and Power based Routing Protocols Routing protocols of this kind mainly leverage routing metrics based on consumed power to perform transmission and/or the perceived/generated interference along a multi-hop path through s Minimum Power Routing As an example, the main objective of the work of Pyo and Hasegawa [17] is to discover minimum weight paths in cognitive wireless ad hoc networks. A detailed system-level picture is presented where the communication system is partitioned into operating system and communication system. A routing weight based on the required power to reach a specific destination is associated with different wireless systems. The proposed routing protocol locally finds the path to minimize the routing weight between a source and a destination. The route discovery procedure is very similar to link state routing algorithms where this newly introduced weight is used. The model does not take into account the PUs, their behavior, or the interference caused by/to other CR nodes. However, such information is implicitly incorporated into routing decisions during neighbor discovery stage. This work introduces a very nicely outlined system model based on multiple interfaces. The performance of the proposed system is highly dependent on the neighbor discovery procedure and its refresh rates as there are no other maintenance or recovery procedures defined in the routing protocol to react to PU activity. Furthermore, the power-level based cost metric is not sufficient to address challenges of multi-hop CRNs Controlled Interference Routing Interference constraints are at the basis of the work in [18] where the authors analyze the tradeoff between single-hop and multi-hop transmission for s constrained by the interference level that PUs can tolerate. Authors analyze the potentialities of a multi-hop relaying by

9 Routing Protocols for Cognitive Radio Networks: A Survey 161 deriving the geometric conditions under which a is admitted into a spectrum occupied by a PU. On the basis of these geometric results authors propose two routing methods termed Nearest-Neighbor Routing (NNR) and Farthest-Neighbor Routing (FNR) Delay-based Routing Protocols The quality of routing protocols can also be measured in terms of delays to establish and maintain multi-hop routes and to send traffic through the very same routes. Besides classical delay components for transmitting information in wireless networks, novel components related to spectrum mobility (channel switching, link switching) should be accounted for in multi-hop CRNs. Delay-aware routing metrics are proposed in [19] [22], which consider different delay components including: a). the Switching Delay that occurs when a node in a path switches from one frequency band to another; b). the Medium Access Delay based on the MAC access schemes used in a given frequency band; c). the Queuing Delay based on the output transmission capacity of a node on a given frequency band Switching, Medium Access, and Queuing Delay based Routings The novelty of work in [19], [20] is the introduction of a metric for multi-hop CRN which is aware of both the switching delay between frequency bands (D switching ) and backoff delay (medium access delay) within a given frequency band (D backoff ). The work of [21], [22] deals with queuing delay. In [22] it is shown that the queuing delay estimation is fairly accurate, and the end-to-end delay provided by the proposed routing protocol outperforms traditional routing protocols Effective Transmission Time (ETT) based Routing A distributed resource management strategy to support video streaming in multi-hop cognitive radio networks is presented in [23]. Given the characteristics of the traffic flows, the main objective is to minimize the end-to-end delay experienced by each video flow based on its

10 162 Journal of Information Industrial Engineering, Vol. 16, August 2010 classes. The authors argue that a centralized protocol would not be realistic in this case since it would require a network-wide mechanism to distribute the necessary information to drive the resource management algorithm. Therefore, a distributed approach is introduced to make local radio resource management decisions on partial information about the network state Tree-based Routing Protocols The original Tree-based Routing (TBR) protocol only works on a single wireless system such as IEEE802.11a or 11b, rather than multiple wireless systems. To address this problem, the author in [24] propose an efficient and practical protocol, called Cognitive Tree-based Routing (CTBR) protocol, which extends and significantly enhances the ability of the known TBR protocol to enable the TBR to handle multiple wireless systems. To adapt the cognitive environment, they introduce a new cognitive-aware link metric to indicate the link quality, and propose global and local decision schemes for the route calculation, in which the global decision is to select the route with the best global end-to-end metric, whereas the local decision is for that a forwarding cognitive terminal selects a interface with the least load. A Spectrum-Tree base On-Demand routing protocol (STOD-RP) is proposed in [25] which simplifies the collaboration between spectrum decision and route selection by establishing a spectrum-tree in each spectrum band. The formation of the spectrum-tree addresses the cooperation between spectrum decision and route selection in an efficient way. The routing algorithm combines tree-based proactive routing and on-demand route discovery. Moreover, a new route metric which considers both CR user s QoS requirements and PU activities is proposed. In addition, their work provides a fast and efficient spectrum-adaptive route recovery method for resuming communication in multi-hop CRNs Probabilistic Approaches A routing approach based on a probabilistic estimation of the available capacity of every CR link is proposed in [26]. A probability-based routing metric is introduced; the metric definition relies on the probability distribution of the PU-to- interference at a given over a given channel. This distribution accounts for the activity of PUs and their random deployment. This routing metric is used to determine the most probable path to satisfy a given bandwidth demand D in a scenario with N nodes that operate on a maximum of M orthogonal frequency bands of

11 Routing Protocols for Cognitive Radio Networks: A Survey 163 respective bandwidths W 1,...,W M (in Hz) Throughput-based Routing Protocols Throughput maximization is the main objective of the routing protocols described here Path Spectrum Availability based Routing Throughput maximization by combining end-to-end optimization with the flexibility of link based approaches to address spectrum heterogeneity is proposed in SPEctrum-Aware Routing Protocol (SPEAR) [27], a robust and efficient distributed channel assignment and routing protocol for dynamic spectrum networks based on two principles: integrated spectrum and route discovery for robust multi-hop path formation, and distributed path reservations to minimize inter- and intra-flow interference. Through simulations and test bed measurements, they show that SPEAR establishes robust paths in diverse spectrum conditions and provides near-optimal throughput and end-to-end packet delivery latency. SPEAR performs extremely fast flow setup and teardowns, and can maintain interference-free flows in the presence of variance in channel availability Spectrum Utility based Routing Achieving high throughput efficiency is the main goal of protocol ROSA [28]. Opportunities to transmit are assigned based on the concept of spectrum utility and routes are explored based on the presence of spectrum opportunities with the objective of maximizing the spectrum utility. The proposed routing protocol is further coupled with a cooperative sensing technique which leverages both physical sensing information on spectrum occupancy and virtual information contained in signaling packets exchanged by s. The exchange of additional virtual information is performed through a common control channel and is used by the local spectrum/power allocation algorithm Link Quality/Stability-based Routing Protocols This section overviews proposed routing protocols which shift the focus to designing stable

12 164 Journal of Information Industrial Engineering, Vol. 16, August 2010 and quality multi-hop routes Routings with Enhanced Path Recovery Functionalities The work presented in [29] presents an algorithm for handoff scheduling and routing in multi-hop CRNs. One of the main contributions of this work is the extension of the spectrum handoff to a multi-link case. Following a classical approach, the problem of minimizing latency for spectrum handoff across the network is shown to be NP hard and a centralized and a distributed heuristic algorithm has been developed. The centralized algorithm is based on the computation of the maximum non-conflict link set. With this approach, the algorithm iteratively assigns new channels to links. To address the starvation problem, an aging based prioritization scheme is utilized. The simulation results show that the performances of the distributed and centralized protocols are very close to each other for the tested scenarios in a grid topology. Both algorithms also provide improvements over the cases where the proposed algorithms are not utilized. Unfortunately, it is not clear how close these algorithms approach the optimal solutions. The implementation details of the distributed algorithm have not been laid out in detail Routings Targeting Route Stability The link stability is considered in the paper of Abbagnale et al. [30] where this parameter is associated, in a innovative way, to the overall path connectivity via a mathematical model based on the Laplacian spectrum of graphs. Paths are measured in terms of their degree of connectivity that in a multi-hop CRN is highly influenced by the PUs behavior. The behavior of a PU is modeled by its average activity factor. The authors introduce a novel metric to weight routes (paths) which is able to capture path stability and availability over time. Indeed, the core idea is to assign weights to routes and paths proportionally to the algebraic connectivity of the Laplacian matrix of the connectivity graph abstracting the secondary network. Route stability oriented routing is presented in [31], where a novel definition of route stability is introduced based on the concept of route maintenance cost. The maintenance cost represents the effort needed or penalty paid to maintaining end-to-end connectivity in dynamic multi-hop CRNs. The maintenance of a route may involve link switching and channel switching

13 Routing Protocols for Cognitive Radio Networks: A Survey 165 operations as PUs become active. In the former case, one or more links along the route must be replaced by other ones not interfered with by PUs, whereas in the latter case, the same link can be maintained, but the transmission must be carried over to another spectrum portion. In either case, signaling is required to coordinate with other s, which translates to a cost in terms of consumed power, and service interruption time while switching routes. 6. Conclusions Since available spectrum bands in CRNs with multi-hop communication are different for each hop, spectrum sensing information is required for topology configuration in CRNs. Moreover, a major design choice for routing in CRNs is the collaboration between routing and spectrum decision. If the optimal route from a to another results in interference to the PUs, end-to-end latency or packet losses can be affected along the route due to this interaction. To mitigate the degradation, multiple spectrum interfaces at intermediate nodes may be selected. Therefore, end-to-end route may consist of multiple hops traversing different spectrum bands. Finally, rerouting needs to be performed in a cross-layer fashion. If a link failure occurs due to spectrum mobility, the routing algorithm needs to differentiate this failure from a node failure. Furthermore, intermediate s can perform rerouting by exploiting the spectrum information available from spectrum sensing functionality select better routes method. In this survey, intrinsic properties and current research challenges of the routing protocols for CRNs are presented. We investigate the unique challenges in routing protocols of CRNs. We strongly believe that research in the field of modeling/designing CRNs routing still needs major contributions explicitly endorsing network dynamics and variability, which are distinctive features of the multi-hop CRNs. To this extent, open research issues in the field of models and algorithms for route designs in multi-hop CRNs have been presented below. Energy management in wireless ad hoc networks is of prime importance because of the limited energy availability in the wireless devices. Over time, various nodes will deplete their battery and drop out from the network; hence the network will eventually become partitioned. In such situations energy conservations are essential to prevent network failures by maximizing the battery lifetime. Almost all of the conventional routing protocols in mobile ad hoc networks (MANETs) that are unaware of battery energy establish only a single shortest path route for data transmission. These protocols may however result in a quick depletion of the battery energy of the nodes along the most heavily used routes and hence the single path routes are

14 166 Journal of Information Industrial Engineering, Vol. 16, August 2010 going to be disconnected. So energy efficient issue is one of the most important factors in designing routing protocol for CRNs. Routing protocol which chooses the best effort route between the source-destination pair for multi-hop communication is called single path routing. Single path routing is not the best solution where the network topology is highly dynamic like CRNAs and where the network resources are limited. Multipath routing is introduced as an alternative to single path routing for its potential to address issues such as route failure and recovery, and network congestion. In energy constrained network such as ad hoc networks multipath routing achieves higher performance than single path routing and at the same time, reduce the power consumption of the relay nodes, alleviating the network partitioning problem caused by the energy exhaustion of these nodes. Because of the rising popularity of multimedia and real-time applications in the commercial environment and the ever growing requirements of mission-critical applications in the military arena, a best-effort routing cannot meet all requirements in most situations. Quality of Service (QoS) support in mobile ad hoc networks has become an important area of research. Compared to the demands of traditional data-only applications, these new requirements generally include high bandwidth availability, high packet delivery ratio and low delay. The QoS framework can perfectly coexist with multipath routing protocols, achieving significant improvements on the overall networking performance, especially for multimedia and real-time applications. The discussions provided in this survey strongly advocate spectrum-aware communication protocols that consider the spectrum management functionalities. The cross-layer design requirement necessitates a rethinking of the existing routing protocols developed for CRNs. Many researchers are currently engaged in developing the communication technologies and protocols required for CRNs. However, to ensure efficient spectrum-aware communication, more research is needed along the lines introduced in this survey.

15 Routing Protocols for Cognitive Radio Networks: A Survey 167 References [1] S. Haykin, Cognitive radio: Brain-empowered wireless communications, IEEE J. Sel. Areas Commun., vol. 23, no. 2, pp , Feb [2] I. F. Akyildiz, W. Y. Lee, M. Vuran, and S. Mohanty, Next generation/dynamic spectrum access/cognitive radio wireless networks: A survey, Comput. Netw., vol. 50, no. 13, pp , Sep [3] R. W. Thomas, L. A. DaSilva, and A. B. MacKenzie, Cognitive networks, Proc. IEEE DySPAN, pp , Nov [4] I. F. Akyildiz, W. Y. Lee, M. C. Vuran, and S. Mohanty, A Survey on spectrum management in cognitive radio networks, IEEE Commun. Mag., vol. 46, no. 4, pp , Apr [5] IEEE Working Group. Wireless LAN medium access control (MAC) and physical layer (PHY) specifications, IEEE Standard, [6] P. Kyasanur and N. H. Vaidya, Protocol design challenges for multi-hop dynamic spectrum access networks, Proc. IEEE DySPAN, pp , Nov [7] C. Cormio, K. R. Chowdhury, A survey on MAC protocols for cognitive radio networks, Ad Hoc Netw., vol. 7, no. 7 pp , [8] K. Chowdhury, I. Akyildiz, Cognitive wireless mesh networks with dynamic spectrum access, IEEE J. Sel. Areas Commun., vol. 26, no. 1, pp , [9] I. F. Akyildiz, W.-Y. Lee, K. R. Chowdhury, CRAHNs: Cognitive radio ad hoc networks, Ad Hoc Netw., vol. 7, no. 5, pp , [10] M. Cesana, F. Cuomo, E. Ekici, Routing in cognitive radio networks: Challenges and solutions, Ad Hoc Netw., Article in press, [11] FCC, FCC , Unlicensed operation in the TV broadcast bands, Nov [12] C. Xin, B. Xie, C.-C. Shen, A novel layered graph model for topology formation and routing in dynamic spectrum access networks, Proc. IEEE DySPAN, pp , Nov [13] X. Zhou, L. Lin, J. Wang, X. Zhang, Cross-layer routing design in cognitive radio networks by colored multigraph model, Wirel. Pers.Commun., vol. 49, no. 1, pp , [14] Y. Hou, Y. Shi, H. Sherali, Optimal spectrum sharing for multi-hop software defined radio networks, Proc. IEEE INFOCOM, pp. 1 9, 2007.

16 168 Journal of Information Industrial Engineering, Vol. 16, August 2010 [15] Y. Hou, Y. Shi, H. Sherali, Spectrum sharing for multi-hop networking with cognitive radios, IEEE J. Sel. Areas Commun., vol. 26, no. 1, pp , [16] M. Ma, D. Tsang, Joint spectrum sharing and fair routing in cognitive radio networks, Proc. IEEE CCNC, pp , [17] C.W. Pyo, M. Hasegawa, Minimum weight routing based on a common link control radio for cognitive wireless ad hoc networks, Proc. IWCMC, pp , [18] M. Xie, W. Zhang, K.-K. Wong, A geometric approach to improve spectrum efficiency for cognitive relay networks, IEEE Trans. Wirel. Commun., vol. 9, no. 1, pp , [19] H. Ma, L. Zheng, X. Ma, Y. Luo, Spectrum aware routing for multi-hop cognitive radio networks with a single transceiver, Proc. CrownCom, pp. 1 6, [20] G. Cheng, W. Liu, Y. Li, W. Cheng, Spectrum aware on-demand routing in cognitive radio networks, Proc. IEEE DySPAN, pp , [21] G. Cheng, W. Liu, Y. Li, W. Cheng, Joint on-demand routing and spectrum assignment in cognitive radio networks, Proc. IEEE ICC, pp , [22] Z. Yang, G. Cheng, W. Liu, W. Yuan, W. Cheng, Local coordination based routing and spectrum assignment in multi-hop cognitive radio networks, Mob. Netw. Appl., vol. 13, no. 1-2, pp , [23] H.-P. Shiang, M. van der Schaar, Distributed resource management in multi-hop cognitive radio networks for delay-sensitive transmission, IEEE Trans. Veh. Tech., vol. 58, no. 2, pp , [24] B. Zhang, Y. Takizawa, A. Hasagawa, A. Yamaguchi, and S. Obana, Tree-based routing protocol for cognitive wireless access networks, Proc. IEEE WCNC, pp , [25] G.-M. Zhu, I. Akyildiz, G.-S. Kuo, STOD-RP: A spectrum-tree based on demand routing protocol for multi-hop cognitive radio networks, Proc. IEEE GLOBECOM, pp. 1 5, [26] H. Khalife, S. Ahuja, N. Malouch, M. Krunz, Probabilistic path selection in opportunistic cognitive radio networks, Proc. IEEE GLOBECOM, pp. 1 5, [27] A. Sampath, L. Yang, L. Cao, H. Zheng, B.Y. Zhao, High throughput spectrum-aware routing for cognitive radio based ad-hoc networks, Proc. CrwownCom, 2008.

17 Routing Protocols for Cognitive Radio Networks: A Survey 169 [28] L. Ding, T. Melodia, S. Batalama, M.J. Medley, ROSA: Distributed joint routing and dynamic spectrum allocation in cognitive radio ad hoc networks, Proc. ACM MSWiM, pp , [29] W. Feng, J. Cao, C. Zhang, C. Liu, Joint optimization of spectrum handoff scheduling and routing in multi-hop multi-radio cognitive networks, Proc. IEEE ICDCS, pp , [30] A. Abbagnale, F. Cuomo, Gymkhana: a connectivity-based routing scheme for cognitive radio ad hoc networks, Proc. IEEE INFOCOM, pp. 1 5, [31] I. Filippini, E. Ekici, M. Cesana, Minimum maintenance cost routing in cognitive radio networks, Proc. IEEE MASS, pp , 2009.