Study On Spectrum Access Technology In Cognitive Radio Network

Recently, the increasing development of wireless commnnication technologies and thediversification of customer’s requirement have generated an explosive demand on more radiospectrum, the shortage of spectrum resource becomes serious. However, studies have reportedthat the existing spectrum resource is severely under utilization in both the time and frequencydomains. Cognitive raido (CR), which has emerged as a dynamic spectrum access technology,is designed to balance the tradeoff between the shortage of spectrum and the inefficientutilization, becomes a research focus. In CR, the wireless devices which have licenses to usethe spectrum are referred to as primary users (PUs), and the unlicensed users are referred to assecondary users (SUs). SUs can access the PU channels while the interference to PUs is underthe acceptable thresholds. Thus, the utilization of spectrum is improved. The SUs performspectrum sensing to detect the status of channels. Then, spectrum access technology is usedfor the SUs to share the spectrum with PUs, based on the sensing results. Recently, there aremany studies about spectrum sensing and spectrum access based on different researchfoundations. However, in some special scenarios, there are still many problems aboutspectrum sensing and access. According to the recent research, in this paper, we study thespectrum sensing and access technology, and their application in vehicular networks. Themajor contributions of this paper are shown as following.1We present a generalized spectrum access strategy in a multi-channel CR network. Theexisting spectrum access strategies include: Overlay, Underlay and Mixed accessstrategies. In the multi-channel network, the SU cannot obtain larger throughput in theexisting strategies. In the proposed generalized access strategy, the SU selects part ofchannels for spectrum sensing, and accesses these channels via mixed access strategy.Moreover, it accesses the other channels directly via underlay access strategy. The optimalstopping theory is used to find the optimal sensing channels and the sensing timedurations of the selected channels are obtained via convex optimization theory. We provethat not all channels are needed for spectrum sensing, the SU can obtain larger throughput.The optimal sensing channel number exists and is unique. Illustrative results indicate theefficiency of the proposed access strategy and algorithm. 2We present a sensing time and transmission power joint optimization algorithm in anuplink model. When an SU system shares multi-channel with PUs, how to select sensingchannel and how to adjust the transmission power after spectrum sensing, are researchfocus in this field. The SU system includes a BS and some SUs. The BS detects the PUchannels, and the SUs adjust their transmission powers based on the sensing results. Thesensing time of BS and the transmission power of SUs are jointly optimized, in order tomaximize the throughput of SU system. According to the theory analysis, we find that theBS doesn’t need to sense all channels. Thus, we propose two sensing channel selectionalgorithms: rate-based and QoS-based sensing channel selection algorithm. Simulationresults show that the proposed algorithm can efficiently improve the throughput of SUsystem.3We present a channel selection algorithm based on Partially Observable Markov DecisionProcess (POMDP) in multi-slot nulti-channel scenario. In a dynamic spectrumenvironment, POMDP is used to express the actions of SU system. At the beginning ofeach slot, the SU should determine1) selects which channels for spectrum sensing;2)how long is the sensing time and3) the transmission power of each channel after sensing.Thus, an optimal policy and myopic policy are proposed to solve the POMDP problem. Inaddition, we present a sensing time and transmission time joint optimization algorithm. Inthe single channel case, the transmission time has two values based on the differentsensing results, under the mixed access strategy. A convex optimization problem isproposed, considering the interference to PUs and the spectrum lost of SUs. Moreover, inthe multi-channel case, the SU detects the channel status via wideband spectrum sensing,and accesses the channels via mixed access strategy. The sensing time and transmissiontime are jointly optimized to maximize the aggregate throughput of SU system.Simulation results show that the proposed joint optimization algorithms outperform thefixed parameters algorithm.4We study the application of CR in vehicular network. Considering the interference tothroughput from path loss exponent of the channel, we present an energy efficientspectrum access algorithm in cognitive vehicular ad hoc network. When the vehicle isclosed to the Access Point (AP), it transmits data to AP. CR is used to find the available spectrum. In different locations, the vehicle selects different transmission strategiesadaptively (direct transmission or relay transmission). When the distance between vehicleand AP is long, the vehicle sends data to AP via a relay, direct transmission is not efficient,for the reason of path loss. However, when the distance becomes short, the vehicle cansend data to AP directly. The switching time between two strategies, and the transmissionpower of vehicles are jointly optimized, in order to minimize the total energy consumption,while maintaining the Quality of Service (QoS) requirement of vehicles. Numerical resultsshow the efficiency of the proposed algorithm.Above all, the spectrum sensing and access algorithms are studied in this paper, especiallyin the model that a single SU system shares multi-channel with PUs system. Moreover, theapplications of CR in vehicular network are studied. Numerical results illustrate the efficiencyof the proposed algorithms.