Signal processing for cooperative cognitive radio networks
Date of Issue2017-03-06
School of Electrical and Electronic Engineering
In recent decades, cognitive radio has been introduced as a practical solution to solve the frequency scarcity problem by setting up a secondary network besides the primary one through a cognitive radio network (CRN) to increase the utilization rate of licensed frequency bands. There exist many parameters to achieve a reliable and practical CRN and it is necessary to conduct extensive signal processing studies to optimize them. It is worth mentioning that the protection of primary signals, belonging to licensed users, is a crucial requirement in establishing of a CRN. On the other hand, conducting a strict sensing method may severely reduce the chance of a secondary network to utilize a free spectrum channel, where there is no point to using a CRN anymore. Therefore, modeling a spectrum sensing mechanism complying both networks' requirements is a substantial part of the CRN design. Since spectrum sensing is a sensitive and important part of the CRN design from the viewpoint of signal processing, it has become the matter of interest of advanced research activities. There are many approaches to optimize the detection performance, and it is worthwhile to note that some proposed methods add new parameters to the system and we need to take them into account. For example, cooperative spectrum sensing (CSS), which is suggested to solve the hidden terminal problem, entails two important parameters as the quality of reporting channels (links between secondary users and fusion center) and the number of cooperative users. Moreover, presuming the possible usage of the CRN concepts in new wireless technologies such as device-to-device (D2D) communication over Long Term Evolution Advanced (LTE-A) standard, it motivates us to think beyond the traditional applications of the CRN by considering some other parameters such as randomness in the number of available secondary users and higher traffic rate of the primary users as compared with the wireless regional area network (WRAN - IEEE 802.22). In this dissertation, we focus on some possible complexities which may be faced when the CSS has been utilized. By considering them step-by-step, we try to reach a more realistic system model which makes our analyzed CRN more feasible. In each step, some performance parameters of the system (such as detection reliability and transmission throughput) would be studied based on several proposed methods and new ideas will be introduced to improve sensing reliability and transmission throughput of the system. “Sensing channels”, “reporting channels”, “number of cooperative users” and “traffic rate of the primary network” are four important parameters to be focused on in this research. Note that many studies have been done under simple assumptions such as AWGN sensing channel, perfect reporting channels, deterministic number of available sensing users and a low-traffic rate of the primary network. Although they could represent a real-life case, focusing on the desired performance criterion optimization over Rayleigh-faded sensing and erroneous reporting channels is still the primary interest of this research work. For the first step, we suppose that the primary network has a low traffic rate and the number of available cooperative users is deterministic. We aim to analyze some pre-existing methods such as improved energy detector (IED) or double-threshold comparison scheme in terms of detection performance and transmission throughput when a CSS-based detection scheme has been implemented among the secondary users. We seek to study their efficiency, especially in the case of imperfect reporting channels. Furthermore, to meet both reliability (protection of the primary signal) and spectrum utilization requirements for throughput analysis, the throughput of the secondary network is maximized over some required constraints. We then extend our analyses to the case of medium-traffic primary networks. Although the analyzed period might be very short, it is most likely that the status of the primary user alters during the duration. This situation affects performance of the system due to higher uncertainty on the activity of the primary user. In the final stage, by using stochastic geometry concepts, we would like to take the effect of randomness in distribution of cooperative users into account. Since the number of available users may not be fixed in a practical wireless network, a stochastic model is considered for distribution of the secondary users besides prior assumptions. This is more critical when a cooperative sensing scheme has been utilized among the secondary users. We finally propose our method to enhance the total throughput of the system when it suffers from an unstable primary network, due to a medium-traffic rate, and a random density of the secondary users, due to a stochastic-modeled secondary network, at the same time. By defining the expression for possible interference and required formulation, we demonstrate the advantages of our proposed method in terms of maximum achievable total throughput of the system in the introduced CRN model. It is worth mentioning that both reliability and spectrum utilization are implicitly considered by assuming the total throughput of the system as the desired performance parameter.