Key Technologies of MAC layer in Cognitive Radio Networks (1)

Cognitive Radio (CR) technology effectively relieves the conflict between the lack of spectrum resources and the increasing demand for wireless access through secondary exploitation of the authorized spectrum, people are getting more and more attention.

To enable CR users to use spectrum holes and avoid harmful interference to authorized users, the Media Access Control (MAC) layer of the CR network must not only provide traditional services, for example, media access control and robust data transmission can also support a new set of functions, that is, the effective use of opportunity spectrum without interfering with authorized users. These new functions are embodied in various aspects such as spectrum detection management, access control, dynamic spectrum allocation, security mechanism, and cross-layer design on the MAC layer. The above technologies will be discussed below.

1. MAC layer spectrum detection Management

MAC layer spectrum detection management is mainly used to control the execution of spectrum detection algorithms in the physical layer, such as determining which channels to detect and when to detect. At present, the research of MAC layer spectrum detection management focuses on the selection and optimization of detection policies and detection parameters, including detection mode selection, detection cycle and detection duration settings, detection channel selection, and detection silence period settings.

Based on the different detection time of CR users, detection modes can be divided into periodic detection and On-Demand detection. Periodic detection refers to a certain period of time for CR users to detect channels, not only when data is sent. This method periodically collects the channel status information, which helps to estimate the channel status and quickly locate the spectrum holes. On-demand detection is performed only when the CR user has data to send. Compared with periodic detection, on-demand detection reduces unnecessary overhead, but it takes a long time to detect spectrum holes. According to the "Energy Efficiency" principle, the adaptive selection of the detection mode is realized by compromising the energy consumption of the detection and the latency required to find idle channels.

In periodic detection, it is critical to select an appropriate detection cycle. If the detection cycle is too large, some idle spectrum cannot be detected and some access opportunities are lost, at the same time, it may cause harmful interference due to the failure to detect authorized users in a timely manner. If the detection cycle is too small, it will lead to too frequent detection and consume unnecessary energy. Periodic detection usually includes two mechanisms: synchronous and asynchronous periodic detection. The synchronous periodic detection mechanism sets the same detection start time and the same detection cycle for all channels, which is easy to implement and lacks flexibility. Correspondingly, the asynchronous periodic detection mechanism has received more attention.

In this paper, [1], an adaptive asynchronous detection cycle optimization algorithm is proposed to minimize access loss. The algorithm automatically sets the detection period for each channel, which plays a role in reducing the idle spectrum duration and maximizing access opportunities. However, for each channel, the detection cycle is still fixed, that is, once the optimal detection cycle is selected, it will not change. In essence, it is still a detection mechanism based on a fixed detection cycle.

As a promotion of the fixed-cycle detection mechanism,] a detection mechanism based on variable detection cycle (FSP) is proposed, introducing the "Detection interval control factor ", by adjusting this factor, the detection cycle can be reduced in areas where the channel status may change, which improves the detection efficiency and reflects the flexibility of periodic changes. In order to further promote the variable detection cycle of the FSP mechanism to the random detection cycle, a random detection mechanism (RAPSS) is introduced for detection at random intervals ), A more general detection cycle optimization model (MRM-SPO) is proposed. At the same time, the existence of detection errors caused by the limitations of the Physical Layer Detection Algorithm in practical application is considered, and the impact of CR user delay occupation channel mechanism introduced to avoid collision with authorized user spectrum on detection cycle optimization.

The detection duration is another major parameter for periodic detection. Whether or not it is set properly is essentially an embodiment of the detection quality and detection speed. Shortening the detection duration will lead to a reduction in the detection quality. Increasing the detection duration can improve the detection quality, but at the same time reduce the utilization of available idle spectrum. In addition to the underlying hardware devices and physical layer detection algorithms, the detection duration can also be selected and optimized based on the CR user's idle spectrum utilization, detection speed, and detection performance compromise.

In order to quickly find spectrum opportunities, in addition to optimizing the detection duration, it also involves the selection of detection channels. The existing research on selecting detection channels mainly includes the following aspects: Selecting channels that are most likely to be idle for detection and optimizing the channel detection sequence.

In addition, the setting of detection silence period is also an important research area of MAC layer spectrum detection and management. From the perspective of CR network, when a CR user detects a spectrum in a CR network, it takes some time for other users in the system to remain silent when working on the spectrum, to ensure that the communications between CR users do not interfere with the detection of authorized user signals, this period is called the silence period.

The silent period can be divided into two types based on different implementation methods: Synchronous silence period and asynchronous silence period. In the synchronous silent period mode, the CR user stops transmitting signals on all available channels of the system at the same time. The setting is simple, and each CR user can detect all channels; the asynchronous silent period method means that one or more CR users stop transmitting signals on the specific channel they are using. Each channel may have different silent periods, CR users can only detect the channels they occupy. Because the implementation of synchronous silence period requires the support of dynamic broadband filters, the current research focuses more on Asynchronous silence period, mainly including asynchronous silence period using protective interval and asynchronous silence period without overlapping time.