A multi-channel cognitive radio is a powerful solution for recovering the exhaustion of frequency spectrum resources. In a cognitive radio, although master and slave terminals (which construct a communication link) have the freedom to access arbitrary channels, access channel mismatch is caused. A rendezvous scheme based on frequency hopping can compensate for this mismatch by exchanging control signals through a selected channel in accordance with a certain rule. However, conventional frequency hopping schemes do not consider an access protocol of both control signals in the rendezvous scheme and the signal caused by channel access from other systems. Further, they do not consider an information sharing method to reach a consensus between the master and slave terminals. This paper proposes a modified rendezvous scheme based on learning-based channel occupancy rate (COR) estimation and describes a specific channel-access rule in the slave terminal. On the basis of this rule, the master estimates a channel selected by the slave by considering the average COR of the other systems. Since the master can narrow down the number of channels, a fast rendezvous scheme with a few control signals is established.

Dynamic spectrum access (DSA) exploits vacant frequency resources via distributed wireless access. The two nodes of DSA, master and slave, access different channels, and thus, cannot communicate with each other. To compensate for the access channel mismatch between the two nodes, a rendezvous channel, which exchanges control signals between two nodes, has been considered. The rendezvous channel based on channel-occupancy ratio (COR) adaptively constructs the channel in accordance with the channel occupancy of other systems, and both a high-speed rendezvous channel and high usage efficiency of the frequency resource are accomplished owing to exploitation of the vacant channel. In the rendezvous channel based on COR, the master and slave recognize the channel with minimum measured COR as the superior channel. As the master sends the control signals through the superior channel recognized by the master, the slave accesses to the superior channel recognized by the slave with higher access rate than to the other channels. As a result, the slave can receive the control signals with highly probability and thus high speed rendezvous channel is achieved. If the master and the slave recognize the different channel as the superior channel, the access rate to the other channel should be larger. This is because the slave obtains the opportunity of receiving the control signals through the different channel from the superior channel recognized by slave and thus the high probability that the slave can receive the control signals is maintained. Therefore, the access rate of slave should be constructed in accordance with the recognition of superior channel by master and slave. In this paper, the access rate of slave to the superior channel is optimally constructed using the analyzed probability of completion of rendezvous channel. The analysis of the probability of completion of rendezvous channel includes the recognition of superior channel by master and slave. Even if the master and the slave recognize the different channel, the constructed access rate of slave can maintain the high speed rendezvous channel. From the theoretical analysis and computer simulation, the rendezvous channel based on COR with the optimal access rate to the channel with the lowest COR achieves reduced time for the rendezvous channel.

Recently, broadband wireless communication has been significantly enhanced; thus, frequency spectrum scarcity has become an extremely serious problem. Spatial frequency reuse based on spectrum databases has attracted significant attention. The spectrum database collects wireless environment information, such as the radio signal strength indicator (RSSI), estimates the propagation coefficient for the propagation loss and shadow effect, and finds a vacant area where the secondary system uses the frequency spectrum without harmful interference to the primary system. Wireless sensor networks are required to collect the RSSI from a radio environmental monitor. However, a large number of RSSI values should be gathered because numerous sensors are spread over the wireless environment. In this study, a data compression technique based on spatial features, such as buildings and houses, is proposed. Using computer simulation and experimental evaluation, we confirm that the proposed compression method successfully reduces the size of the RSSI and restores the original RSSI in the recovery process.