Spectrum Management Functions Of Cognitive Radio Networks Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Abstract-The current assignment of wireless spectrum is based on a static assignment of spectrum to the license users. The license users do not utilize the spectrum all the time. In this way large amount of spectrum is not used efficiently and goes unused. The spectrum is a very limited and the inefficiency in the spectrum usage gives birth to a new communication paradigm i.e. to access the limited spectrum opportunistically. This is referred to as Next Generation Networks and Cognitive Radio (CR) Networks. The novel Spectrum Management functionalities of CRN are explained and open research issues are discussed.

Keywords-Cognitive Radio Networks, Functions of Cognitive Radio Networks, Dynamic Spectrum Access, Next Generation Networks.


The common approach to spectrum allocation is based on statically allocating long-term licenses on portions of the spectrum to providers and their users by government agencies. This type of spectrum allocation leads to underutilization of spectrum as the number of users to access the spectrum has increased dramatically over the years. To address this problem FCC has approved the usage of unlicensed devices in licensed band. Therefore, the option of reusing assigned spectrum when it is temporarily (and locally) available is frequently referred to as Cognitive Radio Network (CRN) which tries to increase the efficiency of spectrum usage. The key enabling technology of CRN networks is cognitive radio (CR) technology, which provides the capability to share the wireless channel with licensed users in an opportunistic manner. A ''Cognitive Radio'' is an intelligent radio which changes its transmission parameters based on the requirements and channel conditions after interacting with the operating environment [1]. The concept of CRN is to identify spectrum holes or white spaces and use them to communicate as shown in Fig 1. White spaces change over time, a cognitive radio is used to "jump" from one chunk of spectrum to another. The sensing capability of the RF front-end over a large range of frequencies in CR transceivers provides the novelity.

There are unique challenges imposed in CRN due to the high fluctuation in the available spectrum and different quality of service (QoS) requirements. In order to address these challenges, each CR user in the CR network must:

• Determine the portions of spectrum available

• Select the best available channel

• Coordination of access to other users about a channel

• On detection of Primary user switching of channel

Fig. 1 [1] Spectrum Hole

These capabilities can be realized through spectrum management functions that address four main challenges: spectrum sensing, spectrum decision, spectrum sharing, and spectrum mobility. The CR network communication components and their interactions are shown in Fig. 2. From the number of interactions it is evident that the CR network functionalities require a cross-layer design approach. Cooperation of spectrum sensing and sharing is required for better spectrum efficiency. Keeping in view the dynamic nature of the underlying spectrum, the upper layers also cooperate with each other in spectrum management and spectrum mobility functions.

This paper presents the definition, architecture and functions of the CRN. In sec 2 we will present the architecture of the CRN Networks. Spectrum Management Functions and their challenges are discussed in section 3. In section 4 we will discuss the upper layer issues and its challenges. Finally, we will conclude the paper in section 5.

Fig. 2 Spectrum Management Framework

CRN network architecture

The components of CRN network architecture are shown in the figure 3(a) and all possible scenarios are considered.

Primary Network: A spectrum band or licensed network which is owned only by the licensed users. The primary network consists of a primary user and a primary base station.

Next Generation Network / CRN Network: No license is required for any band. However, spectrum is opportunistically accessed. They can be of both the types' i.e. Infrastructure network and an adhoc network as shown.

The components of a CRN network are as follows:

CRN Network user (unlicensed user, secondary user)

CRN Network Base Station.

Spectrum Broker shares the spectrum resources among different CRN networks users.

Fig. 3a[1] Architecture of CRN

CRN network Functions

The CRN Network has the capability to operate in both licensed band as well as in unlicensed band.

In a licensed band the CRN user will first sense a spectrum hole with the help of cognitive communication techniques in the licensed band. In the licensed band the CRN functions are mainly aimed at the interference avoidance with the primary users. Different techniques have been proposed which are discussed in detail in the latter section for spectrum sensing. The most important issue is to avoid interference with the primary user in this band. As soon a primary user is detected the CR user should vacate the band and find a new band for communication, this is called spectrum handoff. All network entities have equal rights to access the spectrum as there are no license users in the band.

Spectrum Sensing

A cognitive radio is designed to adapt to the changes in the spectrum band which makes spectrum sensing the most important Cognitive radio function. Spectrum Aware protocols are required to cater with the changing conditions of spectrum. Following are the main functions that spectrum sensing provides as shown in fig 3b:

Fig. 3b Spectrum Sensing Structure

Primary User Detection: CRN users sense the spectrum and analyse it by their local observation. Efficient way of spectrum holes detection is to detect the primary users which are receiving data within the range of a CRN user [1]. Thus PU and Spectrum holes both are identified.

Cooperation: Cooperation of spectrum sensing and sharing is required for better spectrum efficiency which is done by sharing information between different CRN users.

Sensing Control:

This allows the user to adapt to the dynamic nature of radio environment.

The main purpose of spectrum sensing is to find spectrum holes to provide more spectrum access opportunities to CRN network users without interfering with the primary network user. The current technology of radio frequency (RF) front-ends is unable to sense and transmit at the same time, which decreases the opportunities to transmit leading to sensing efficiency problem [2].

Since cognitive radio (CR) networks are responsible for detecting the transmission of primary networks and avoiding interference to them, CRN networks should intelligently sense the primary band to avoid missing the transmission of primary users. Thus, sensing accuracy has been considered as the most important factor to determine the performance of CR networks. Hence, recent research has been focused on improving the sensing accuracy for interference avoidance [2].

For the detection of spectrum holes different techniques have been developed. We will go through these techniques in the following subsection Detection and then in the next section we will discuss the open research issues in this area.

Detection: In [1], three spectrum sensing methods are discussed, transmitter detection, cooperative detection and interference based detection.

Fig 3c Classification of Sensing

Transmitter Based detection: This approach is basically based on the local observation of a CRN user. Weak signals are detected from the primary transmitter by the CRN user as shown in the Fig. 6.

A hypothesis model is used in transmitter based detection for sensing

r(t) is the CRN user's received signal, s(t) is the signal transmitted by the PU transmitter, n(t) is noise (AWGN) and amplitude gain is given by h. There is no presence of licensed user signal corresponds to H0 hypothesis and alternately corresponds to H1 in presence of PU signal in a band. Transmitter based detection schemes are as follows:

Match Filter detection: Match Filter Detection has a fast sensing time. The most important thing about match filter detection is very tight synchronization between transmitter and user is required & prior knowledge about the primary user signal is also required. The performance of match filter is greatly degraded in case of inaccurate information. Cost & complexity rises as a number of matched filters are required for different PU signals [1].

Fig. 4a Match Filter Detection

Energy Based Detection: This is optimal technique in case of known noise in the system. The presence of a primary user is detected as shown in the figure 4b. The received signal is first squared and then integrated over the interval and finally measured against the threshold to find the presence of primary user over the interval T. If there is uncertainty in noise power this technique's performance is highly degraded. Signal types cannot be differentiated in this technique so a tight synchronization is required between all the CRN users for sensing and transmission to avoid false alarms.

Fig 4b Block Diagram of Energy Based detection

Feature detection: In this technique the primary user is detected by exploiting the cyclo stationary characteristics or features of the received signal such as cyclic prefix or modulation types etc. This technique works really well in case of unknown noise but is computationally complex and requires a longer time for sensing. The differentiation of noise from modulated signal is carried out by the spectral correlation function which is the main advantage [3]. The primary user is detected by averaging the received signal r (t) over the interval T and comparing the result with the test stats.

Fig 4c. Block Diagram of Feature Based Detection

Primary Receiver Detection

The primary receiver detection is an efficient way of PU detection. The detection is carried out by exploiting the primary receiver's local oscillator leak power which is emitted by the RF front end as shown if fig 5.

Fig. 5. Receiver Based Detection

Cooperative detection

In most cases a PU network and CRN are physically separated so they do not interact but in transmitter based detection due to lack of primary user information interference occurs. This is because the observation range is less as compared to transmission range and as a CR user finds the spectrum available for transmission it may interfere with the primary user which is called the receiver uncertainty problem. Cooperative detection shares the sensing information from other users for more accurate detection. Cooperative detection technique supports both the centralized or ad-hoc architectures. While cooperative approach provides more accurate sensing they impose overhead traffic.. The multi-path fading and shadowing effects can be reduced so that the detection probability is improved in a heavily shadowed environment [10].

Interference based detection

A new model introduced by the FCC called the interference temperature model [1]. The idea is that by taking a single measurement, a cognitive radio can completely characterize both interference and noise with a single number. Interference and noise behave differently, interference is typically more deterministic and uncorrelated to bandwidth, whereas noise is not. The interference temperature limits are defined by the FCC for a given geographical region. This value would be a maximum amount of tolerable interference for a given frequency band in a particular location and as long as CRN user does not exceed this level it is allowed to carry its operation in the band. The CR users are not able to distinguish between the PU signal and noise so this technique is not used and most of the concentration is based on transmitter based detection.

Although all these efforts enable CR users to enhance the sensing accuracy, but the current hardware limitations of CR users should monitor the spectrum continuously through the radio frequency (RF) front-end to avoid interference, [2]. But practically the RF front-end cannot differentiate between the signals i.e. PU/ CRN signal. [8]. for this reason CR users necessitate a periodic sensing structure where sensing and transmission operations are performed in a periodic manner with separate observation period and transmission period. In this structure, CR users should stop their transmissions during the sensing time to prevent false alarms triggered by unintended CR signals.

This periodic sensing structure introduces the Interference avoidance & Sensing efficiency

There is a trade-off between interference and sensing efficiency. For interference avoidance, the observation time needs to be long enough to achieve sufficient detection accuracy, i.e., longer observation time leads to higher sensing accuracy, and hence to less interference. But as the observation time becomes longer, the transmission time of CR users will be decreased. Conversely, while a longer transmission time enhances the sensing efficiency, it causes higher interference due to the lack of sensing information. Hence, observation time and transmission time are the sensing parameters that mainly influence both the spectrum efficiency and interference avoidance. Thus, the proper selection of these sensing parameters is the most critical factor influencing the performance of CR networks [2].

Spectrum sensing challenges

A number of open research challenges for spectrum sensing exists which require developments. Some of which are discussed below:

Detection capability: Development of novel spectrum sensing algorithms is required for reducing the sensing time and increasing the spectrum accuracy.

Optimization of cooperative sensing: Cooperative sensing improves the sensing accuracy by requesting information by other CR users but this also increases the traffic over the network which increases the latency. These factors need consideration for better and accurate sensing.

Spectrum decision

Spectrum holes are available over a large range of frequencies which include both the licensed and unlicensed band. Now theses spectrum holes which are detected after spectrum sensing show different characteristics. Spectrum Decision should perform the following tasks before choosing a band for transmission.

Characterization of all available Spectrum Bands: The characterization of spectrum is basically the analysis of the radio environment and the PU activity. The characteristics of spectrum holes change over time so each hole should be characterized by the changing radio environment as well as other parameters which are required power for transmission, frequency of operation, bandwidth available, wireless link errors and link layer delays etc. Different models have been proposed to estimate the statistical behaviour of PU in a primary band which is discussed in…….

Selection of a specific band: Once all the spectrum bands are characterized then appropriate band according to the QOS should be selected to carry the transmission. As we know the entire communication session process may consist of multiple hops over heterogeneous spectrums band selection should coordinate with the routing protocols to carry end to end session of communication.

Reconfiguration: Now after selection of the appropriate band according to the requirement and route selection, reconfiguration is involved. Reconfiguration helps to adapt to the channel characteristics which change over time.

Spectrum Decision Challenges

Cooperation with reconfiguration: After the characterization of a spectrum band the band should be chosen which meet the QOS requirement of the CRN user. However the channel characteristics change over the time so reconfiguration is required there to meet the QOS requirement of the CRN user. Therefore a framework is required to carry jointly the spectrum decision and reconfiguration [1].

Spectrum decision over heterogeneous spectrum bands: The CRN users will have an opportunity to operate in both the unlicensed bands as well as licensed bands. In unlicensed band the CRN users and the other users all have the equal rights to access and operate the spectrum while in licensed bands the CRN users do not hold the exclusive right to operate it should consider the PU activity and carry its transmission only in the absence of PU. So to decide between the heterogeneous bands the spectrum decision should support decision over all the bands. Novel techniques are required to choose between all the bands in accordance with the QOS requirement of a CRN user [1].

PU activity models: Novel PU activity models need to be developed to have a better accuracy and lower interference with primary users.

Spectrum Mobility

Spectrum mobility is the process when a CRN user changes its frequency of operations. The target of a user is to capture the best available spectrum to carry its transmission.

Spectrum handoff

In CRN Networks the operating frequency is changed when there is an appearance of primary user or channel condition does not meet the QOS required for a CRN user. So the spectrum mobility gives rise to the spectrum handoff.

In the spectrum mobility whenever the handoff takes place the operating frequency changes which also changes operating protocols in the network to another mode. The transition should be made smooth and as soon as possible so that the application should not suffer from degradation of performance.

Spectrum Mobility Challenges

There are a number of frequency bands available, so to choose between the best bands based on its characteristics and CRN user requirement is a challenging task and novel techniques are required for this purpose.

The next challenge is to minimize the delay and loss during the transition of CRN user when a spectrum is chosen.

Novel algorithms are required for the ensured of less performance degradation during transition.

As the available channels change over time the quality of service maintenance is challenging.

If a user moves from a place to another, the continuous spectrum allocation is a major challenge in CRN networks.

Spectrum Sharing

In [11], spectrum sharing is defined as when two systems accessing the same spectrum the primary user in CRN network can share the same spectrum with a CRN user. One example is such a scenario that a cellular provider leasing its unused spectrum to a CRN user when cellular traffic is expected to be significantly lower .The spectrum sharing with primary user and selecting frequency from a wide range of available spectrum are the main reasons of different challenges in open spectrum usage.

There are five major steps in the spectrum sharing process [1].

Fig 6 Spectrum Sharing Techniques

Fig 7 Inter-Network & Intra- network in CRN

Spectrum sensing: The solution and the challenges for spectrum sensing are discussed above in this paper.

Spectrum allocation: The channel is allocated based on spectrum availability and on the policy of not interfering with the primary network user. So designing a novel spectrum policy for the improvement of a node is a hot research area.

Spectrum access: A major problem in spectrum access is the collision by multiple users in overlapping portion of the spectrum. To prevent this collision a solution to this problem is required.

Transmitter-receiver handshake: The transmission-receiver handshake protocol is important to inform the receiver of this communication when a portion of spectrum is determined.

Spectrum mobility: Spectrum mobility is very important for successful communication when a specified portion in use by CRN user is acquired by the licensed holder or user.

Spectrum sharing Techniques

The spectrum sharing techniques are classified and the fundamental results about these techniques in CRN networks are described here. The solutions for spectrum sharing in CRN networks are classified in three aspects [1] based on the architecture, access behavior and access technology as shown in Fig 6.

Architectural Based Techniques:

Centralized spectrum sharing: Spectrum allocations and the access procedures are controlled by a centralized entity in these solutions [1,12]. Here a central entity in the CRN network constructs a spectrum allocation map after receiving the spectrum allocation measurements from each entity of the CRN network.

Distributed spectrum sharing: In distributed spectrum sharing there is no existence of a central entity so the proposed solutions are distributed solutions [11]. Here every node in the CRN network is responsible for the spectrum allocation.

Access Behaviour Techniques:

The access behaviour based technique, which can be cooperative or non-cooperative, is explained below.

Cooperative spectrum sharing: The effect of the communication of a node on other nodes is the main consideration of these solution techniques [1]. All the information are made available to all nodes which are collected by every individual node.

Non-cooperative spectrum sharing: Only the node at hand is the main consideration of non-cooperative solution also known as selfish solution [11]. This selfish solution may reduce the utilization of spectrum.

Cooperative spectrum sensing approach enhances the accuracy and simulations have also been performed between cooperative and non cooperative approaches proving cooperative approaches as a better solution [1].

Spectrum Access Techniques:

The spectrum access technique is based on the technology and is explained below.

Overlay spectrum sharing: In this technique the available portion of the spectrum which is not used by primary users is accessed by a CRN user which minimizes the interference to the primary system [13].

Underlay spectrum sharing: This technique acquires a spectrum allocation map and then selects portions of a spectrum. A CRN node transmits power at that portion of a spectrum which is considered as noise at the PU. This technique, when compared to overlay technique, can use more bandwidth [1].

Existing Spectrum Sharing Techniques

Inter-network spectrum sharing and intra-network spectrum sharing are the existing spectrum sharing techniques which are the combination of the above discussed classifications as shown in Fig 7.

Inter-Network Spectrum Sharing:

Some unique challenges in CRN networks are posed in inter-network spectrum sharing.

Centralized Inter-Network Spectrum Sharing

In [13], the Common Spectrum Coordination (CSCC) protocol is proposed for the co-existence of both IEEE 802.11b and 802.16a networks. This protocol necessitates modification in using both the networks therefore we do not consider this as a complete solution. The assumption in this is that each node is has a cognitive radio, narrow band control radio and a low-bit rate. Then each node broadcasts the CSCC message with other node and so the co-existence is maintained.

Distributed Inter-Network Spectrum Sharing

D-QDCR (Distributed QoS based dynamic spectrum reservation) is a scheme in which a BS with high priority allocates a spectrum to its user according to the requirements of the user. Here a basic unit Q-frame is allocated by the BS and then control and data channels are used for coordination and communication of data between the users. The D-QDCR scheme serves important contribution in distributed inter-network spectrum sharing [1]. The distributed inter-network spectrum sharing provides broader view of spectrum sharing solution including operator policies for determination of spectrum allocation. A common control channel is required for this solution.

Intra-Network Spectrum Sharing:

Here the users of a CRN network try to access the available spectrum avoiding interfering with the PU.

Cooperative Intra-network Spectrum Sharing

In [14], a cooperative local bargaining approach is proposed in which groups formed called barging groups. Minimum spectrum is allocated to each user considering the fairness to access the spectrum. Experimental results state that the complexity is reduced to 50 percent by using cooperative sharing. In [15], a distributed coordination scheme is proposed which also considers local groups. In this scheme it is shown that CR users are equipped with minimum number of Common channels while in cooperative local bargaining scheme the common channel may not exist in CRN networks or can be occupied by primary user.

Non-cooperative intra-network spectrum sharing

A device-centric spectrum management scheme (DCSM) is proposed in [16] where users act independently according to preset spectrum rules when local interference patterns are observed. Here five rules are provided that tradeoff performance with implementation complexity and communication costs. Experimental results show that the proposed rule-based approach reduces communication costs from efficient collaborative approaches by a factor of 3-4 while providing good performance [16].

Spectrum Sharing Challenges

There are many open research issues for the seamless open spectrum operation. In [1], some challenges in spectrum sharing in CRN networks along with some possible solutions are described which are.

Common Control Channel:

CCC facilitates transmitter receiver handshake communication [13] and much other functionality. If a CRN user uses a channel and when a primary user chooses that channel, then the channel has to be made available to the PU without interference. This is also applicable for the CCC. So it is infeasible if fixed CCC is implemented [1] and when CCC is not available, handshake between the transmitter and receiver is challenging. Receiver driven techniques can be exploited in the case where CCC is not used or available.

Dynamic Radio Range:

In CRN networks, the users hop between a large number of frequencies due to which its neighbours change over time and also due to the change in frequency [1], the routing decisions are affected by this change. So a CCC should be carefully chosen considering the dynamic range of radios.

Upper Layer Issues

In CRN networks there are also some upper layer issues related to the network layer and flow and congestion control. Here some challenges are given which are related to these areas.

Routing challenges:

In CRN networks efficient spectrum allocation is a challenge, especially for multi-hops transmissions so novel routing algorithms are required. There are two types of existing solutions based on inter-dependence between selection of route & management of spectrum. In [18], these two design methodologies are investigated which are a decoupled design and a collaborative design. In the decoupled design both tasks are independent and in collaborative design they are integrated into a single task. Experimental results show that the collaborative design performance is better as compared to the decoupled task [18].

The remaining major challenges and open research issues in CRN networks are summarized below:

Common control channel: As we have discussed that the CRN network is lack of a common control channel (CCC) which is a major problem since the discovery of neighbours or route and the route establishment is done by broadcast messages in traditional routing protocols which will be very difficult due to the lack of a CCC [1]. Hence, solutions are required for this problem in CRN networks.

Intermittent connectivity: In CRN networks the available channel can change with the appearance of a primary user in that channel. So models are required for this purpose. Such a channel based model is proposed in [18] but still more models are required for the connectivity purpose in CRN networks [1].

Re-routing: In CRN networks an established route of a CRN user changes as the availability of spectrum or due to the mobility of the user. Hence re-routing algorithms for routing in CRN networks are required [1].

Transport Layer Challenges:

In CRN networks, due to the wireless link errors and access delays the TCP and UDP protocols are affected. Some solutions have been proposed in [19] to reduce the performance degradation of these protocols but still there are some major issues and problems in the transport layer [1].As we know that TCP is dependent on Probability of packet loss & round trip time (RTT) which depends on the frequency in use and on the available bandwidth. And as the frequency and bandwidth may change rapidly so the existing TCP and UDP protocols cannot be used. Therefore novel protocols are required.


The CRN can solve the problem of spectrum underutilization by opportunistically accessing the spectrum and making most out of the limited spectrum. Spectrum scarcity can be minimized by Cognitive Radio Networks. The CR devices need to have spectrum management functions incorporated in it. Spectrum aware protocol need to be developed. In this survey a bottom up approach was followed starting with the Cognitive radios capability to required communication protocols for efficient communication. However for the development of an efficient spectrum aware communication more research is being carried out.