International Core Journal of Engineering 2020-26 | Page 160

2019 International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM) A New PU Perceptual Algorithm for Five Consecutive Sensing Items Fenglin Jing Rui Gao* Zhenghua Zhang Shuo Wang School of Information Engineering Yangzhou University Yangzhou, China [email protected] School of Information Engineering Yangzhou University Yangzhou, China *[email protected] School of Information Engineering Yangzhou University Yangzhou, China [email protected] School of Information Engineering Yangzhou University Yangzhou, China [email protected] advantage of this method is that it does not need priori information and it is not affected by noise uncertainty [8]. However, its shortcomings are also obvious, mainly due to the high complexity of the algorithm. When the signal correlation is weak, the performance of the algorithm decreases sharply. Entropy-based detection is discussed in document [9], its main advantage is that it has better detection performance and is not affected by noise uncertainty. However, its application scope is limited due to many assumptions. As we all know, the Energy detection method [10] is most widely used in practical systems because it does not require a priori information of signals and low computational complexity. Abstract—Energy detector is the most common way of idle spectrum sensing in cognitive radio. However, it is difficult to meet spectrum sensing performance requirements under low SNR conditions. We propose a novel algorithm that considers five consecutive sensing items. The algorithm is named five items energy detection (5IED), which makes a decision in one perception item, taking into account two sensing items immediately before the project and two sensing items immediately after the project. The algorithm proposed in this paper is to use the duty cycle value of the primary user (PU) activity to track the change of PU state. In the presence of noise uncertainty, theoretical analysis and simulation show that 5IED has better detection probability and robustness than classical energy detection (CED) algorithm. At the same time, we compare 5IED with other algorithms in terms of complexity and average detection time. In addition, we also discuss and compare the performance of each detector under different channel models. Therefore, the performance of the new spectrum sensing technology is usually compared to the performance of CED [11]. However, the efficiency of CED detector is significantly reduced in the case of low signal-to-noise ratio (SNR) and noise uncertainty, as occurred in the operating CR scenario [12]. Recently, document [13] has proposed an improved energy detection (3EED) algorithm. The main contribution of this method is that it can greatly improve the detection probability without increasing the complexity of the algorithm. And it has the minimum FAP offset, which can be directly compared with CED. Keywords—Cognitive radio,Energy detection ,Rayleigh fading, Noise uncertainty. I. I NTRODUCTION In cognitive radio (CR), in order to ensure the effective sharing of spectrum, sensor nodes must have fast and accurate spectrum sensing capability, which is an important part of cognitive radio sensor networks (CRSN) [1]. At the same time in the development of 5G communication system [2], it is also a very important link. Spectrum sensing (SS) is of great significance to the implementation of CR systems. The key to spectrum sensing is to detect whether the primary user (PU) signal exists or not, and then use the free spectrum band for opportunistic communication [3]. Inspired by the document [13], in order to solve these problems, we propose a new spectrum sensing algorithm, called the five-items Energy Detection algorithm (5IED), which utilizes the knowledge of the average duty cycle of the PU activity model. In fact, 5IED uses five consecutive sensory items, that is, it makes decisions in one sensory item, taking into account both the previous two items and the subsequent two items. As explained later, when the average duty cycle of PU is low, the 5IED decision threshold expression becomes independent of the parameters of the PU activity model. For the low duty ratio, the analysis shows that 5IED is superior to CED and 3EED. At present, the research on SS at home and abroad includes the following common methods. The advantage of the matched filter method [4] is that the detection time is short and the performance is optimal. However, this requires a priori information [5] of the model and high phase synchronization. Document [6] are cyclostationary detection method which has advantages in distinguishing noise from signal. However, its shortcomings are also obvious mainly due to high computational complexity and long detection time. In [7], a method based on eigenvalue detection is proposed. The 978-1-7281-4691-1/19/$31.00 ©2019 IEEE DOI 10.1109/AIAM48774.2019.00035 The rest of this article is organized as follows. First, in Section 2, we review the advantages and disadvantages of spectrum sensing models and traditional energy detection. Then in Section 3, we detail the process of the 5IED algorithm. In addition, we also derive the false alarm probability and 138