Article Preview
TopIntroduction
In recent years, the number of wireless devices and services are rapidly increasing. However, a lot of licensed wireless spectrum is underutilized by existing wireless services. Many studies are constantly studying to solve the problems that frequency shortage. Cognitive Radio (CR) concept has attracted public attention, as one of the leading technologies to resolving this problem.
The CR proposed by Joseph Mitola in 1999 was based on the fact that most of the radio spectrum was not used. The CR technology has two type of users; one is a Primary User (PU) having a license for the channel and freely access the frequency. The other is Secondary User (SU) which shares a frequency with a PU but only can use when PU is not occupying channel.
Therefore, the CR technology must know precisely about the surrounding the wireless environment. In the CR, the SU is allowed to use the frequency resource only when the PU is not used, and the spectrum hole must be efficiently searched in a wide frequency band. This frequency sharing technology achieves that can reduce the frequency supply constraints and increase the frequency utilization efficiency and consumer convenience. The CR technology has the major functions which are characterized by a cognition cycle of observing, adapting, determining, and using the best opportunity to interact with the radio environment as shown in Figure 1.
In Spectrum Sensing, the CR system detects spectrum holes with the spectrum sensing techniques such as transmitter/energy detection, interference based detection, matched filters and cooperative detection. This sensing step analyzes each channel, and if it finds a PU, it decides to change the channel.
Figure 1. Functional architecture of a CR
In this paper, we construct a small multi-hop CR environment in the lab using Raspberry Pi, Universal Software Radio Peripheral (USRP) and Gnu Radio. We improve existing 1-hop CR architecture such as packet structure, packet forward process. Also, we implement a network layer function on the 1-hop basis architecture to support multi-hop environment. From the experimental result, we present our implemented architecture performance and limitation. Performance result shows our architecture supports multi-hop environment.