In the previous section, primary user detection techniques and their comparison is discussed. With the detection results, secondary users will have an idea on which spectrum bands he/she could use. However, the availability and quality of a spectrum band may change rapidly with time due to primary user's activity and competition from other secondary users. In order to utilize the spectrum resources efficiently, secondary users need to consider issues such as when and how to use a spectrum band, how to co-exist with primary users and other secondary users, and which spectrum band they should sense and access if the current one in use is not available. Therefore, in this section, the existing spectrum allocation and sharing approaches are reviewed. Existing spectrum allocation and sharing schemes can be classified in different types,

• Based on spectrum used by secondary users
• Based on the access technology used by the secondary users
• Based on network architecture
• Based on behavior of secondary users

A) Based on Spectrum Used by Secondary Users: The first classification is according to the spectrum bands that secondary users are using. They can be further divided in to two types depending upon the type of spectrum bands used i.e. Unlicensed and licensed.

i) Open Spectrum Sharing: In this model, all users can access the spectrum equally, subjected to certain constraints on the characteristics of the transmit signal. This approach is used nowadays, in the Industrial, Scientific, and Medical (ISM) unlicensed bands. In this sharing, since no users own spectrum licenses, they all have the same rights in using the unlicensed spectrum. The Wireless Fidelity (WiFi) devices in this bands have both advantages and disadvantages. The free access and ease of approval increased the use of WiFi almost in every device. However, due to the large number of WiFi devices, it has become almost impossible for other services (especially medical and industrial services) to operate in this frequency band with a reasonable quality of service.

ii) Dynamic Exclusive Sharing: In this model, a frequency band is reserved for a particular service exclusively, but different providers can share the spectrum. Sharing results in higher spectral efficiency than the use of $N$ frequency bands by $N$ separate operators. The sharing of the spectrum can be done by a trading (buying/selling or auctioning approach). One more way is that a regulator assigns spectrum to a particular user or service based on time varying basis depending on the usage statistics in a given location. For example, a cell phone service provider might have the right to use 50MHz of spectrum in the morning and only 20MHz at noon.

iii) Hierarchical Access Sharing: In this model different priorities are assigned to different users. Licensed users are given priority, so they are served in such a way that they experience the same service quality as if the spectrum were reserved exclusively for their usage. Non licensed users are allowed to transmit in such a way that they do not affect the performance or service quality of the licensed users. The unlicensed user adaptively decides whether it might uses parts of a spectrum that is assigned by default to licensed users. The key principle of hierarchical access model is used in CR. In other words, the secondary users (unlicensed users) do not disturb the primary users. Such non disturbance can be achieved by two fundamental approaches: overlay, and underlay approach.

B) Based on the Access Technology Used by the Secondary Users: Next classification is based on the licesed spectrum sharing. They can be further divided in two categories,

i) Spectrum Underlay Approach: In spectrum underlay, secondary users are allowed to transmit their data in the licensed spectrum band when primary users are also transmitting. But they keep their transmit Power Spectral Density so low that their interference to primary users is insignificant. In other words, the interference temperature model is imposed on secondary user’s transmission power so that the interference at a primary user’s receiver is within the interference temperature limit and primary users can deliver their packet to the receiver successfully. Spread spectrum techniques are usually adopted by secondary users to fully utilize the wide range of spectrum. However, due to the constraints on transmission power, secondary users can only achieve short-range communication. If primary users transmit data all the time in a constant mode, spectrum underlay approach does not require secondary users to perform spectrum detection to find available spectrum band.

ii) Spectrum Overlay: Spectrum overlay is also referred to as opportunistic spectrum access. Unlike spectrum underlay, secondary users in spectrum overlay will only use the licensed spectrum when primary users are not transmitting, so there is no interference temperature limit imposed on secondary user's transmission. Instead, secondary users need to sense the licensed frequency band and detect the spectrum white space, in order to avoid harmful interference to primary users.

C) Based on Network Architecture: The next classification is according to the network architecture. It is further divided in to two categories depending upon control over allocation technique.

i) Centralized Control: When a central entity controls and coordinates the spectrum allocation and access of secondary users, then the spectrum allocation is referred as centralized.

ii) Distributed Control: If there is no central controller because of the involvement of high cost of constructing an infrastructure or the ad-hoc nature of the network such as for emergency or military use, this kind of spectrum sharing belongs to distributed spectrum sharing. In distributed spectrum sharing, each user makes his/her own decision about his/her spectrum access strategy, mainly based on local observation of the spectrum dynamics.

D) Based on Access Behavior of Secondary Users: The last classification is according to the access behavior of secondary users. It is further divided into cooperative and non cooperative spectrum sharing techniques.

i) Cooperative Spectrum Sharing: If all secondary users work towards a common goal, for example they belong to the same operator or service provider; they will coordinate their allocation and access in order to maximize their social welfare. This is called cooperative spectrum sharing. Most centralized spectrum allocation can be considered as cooperative.

ii) Non Cooperative Spectrum Sharing: It is not always the case that all secondary users belong to the same service provider, such as those who access the open spectrum band. Different users have different objectives, and hence they only aim at maximizing their own benefit from using the spectrum resources. Since users are no longer cooperative in achieving the same objective, this kind of spectrum sharing is a non cooperative one, and secondary users are selfish in that they pursue their own benefit.