Prediction of Channel Availability in Cognitive Radio Networks Using a Logistic Regression Algorithm Hans Marquez #*1 , Cesar Hernández #2 , Diego Giral #3 # Universidad Distrital Francisco José de Caldas, Bogotá-Colombia. * Colciencias, Bogotá-Colombia. 1 hans.marquez.ramos@gmail.com; 1 cahernandezs@udistrital.edu.co; 3 dagiralr@correo.udistrital.edu.co Abstract—The capacity of predicting spectral occupancy in cognitive radio networks offers the possibility of developing better policies in channel assignment to secondary users, according to the predicted spectral opportunities. This work develops a prediction model to determine and exploit spectral opportunities while avoiding the continuous search for channel availability in cognitive radio networks. The proposed scheme creates an availability prediction matrix for every available channel in the GSM band that includes their times of availability. By using this information, there is a potential to improve channel allocation policies. The model contains two processes: the first one performs a training process in order to prepare the prediction algorithm so that it can make more reliable predictions and the second one uses the logistic regression algorithm to estimate the availability in every available frequency which can be profited by secondary users, who intend to start transmissions. Measurements were made for average bandwidth, average delay and prediction error. The results obtained were evaluated with real spectral occupancy data in the GSM frequency band. The developed model shows a low prediction error which enables optimal channel assignment mechanisms, hence minimizing failed handoffs through the channel occupation of primary users. Keyword - Availability, Cognitive radio, Logistic regression, Prediction. I. INTRODUCTION Cognitive radio is defined as a “radio system that knows its environment and can dynamically and autonomously adjust its radio operation parameters” [1]-[3] and offers a solution to the current problem that communications face which is the scacrcity in the available spectrum. However, there is a low use of the spectrum in some bands while some bands are completely saturated. To deal with this issue, the study on cognitive radio network (CRN) began whose advantages would allow a more flexible use of the spectrum therefore optimizing the already limited resources in wireless networks. This would not only avoid the rigidness in the current assignment of the spectrum but would also improve the quality of the service offered resulting in higher spectral efficiency. The generated spectral opportunities allow a non-licensed user, a cognitive radio user or a secondary user (SU) to use a channel from an available licensed band until the primary user (PU) takes over that channel. Other cases include a decrease in the quality of the channel taken by the SU, the interference from a SU with the PcU’s activity and the mobility of the SU leads him to an area out of coverage. Under these considerations, the SU should release this channel and search for a new one which is known a spectral handoff [4]-[6]. The method to avoid or minimize the interference on a PU is to force the SU to perform a handoff or change of channel before the presence/arrival of the PU is detected. This reduces the degradation of the channel quality so that the transmissions of the PU are not affected by the opportunistic use of the spectral resources on behalf of the SU. To generate these handoffs before the PU’s arrival, a model is proposed to predict the spectral occupancy by estimating the possibility of a PU’s arrival which would optimize the channel assignment process. The present work is structured as follows: section 2 presents the related work, section 3 describes the development of the model, section 4 shows the results, and section 5 finishes with a set of conclusions. II. RELATED WORK In [7], the research focuses on how the increasing use of services inside a vehicle (cellphones, GPS and radio signals) has become a concern in terms of security. On that matter, the authors propose a system that detects the driver’s distraction and adapts the vehicle to mitigate such distractions. To develop this system, the use of Support Vector Machines (SVM) was proposed to develop a real time approach for the cognitive distraction using the driver’s ocular movements. The data was collected from a simulation with ten participants that interacted with the system while driving. These data were used to train the SVM and logistic regression models with the purpose of investigating three different characteristics: how was created the distraction, which data ISSN (Print) : 2319-8613 ISSN (Online) : 0975-4024 Hans Marquez et al. / International Journal of Engineering and Technology (IJET) DOI: 10.21817/ijet/2017/v9i5/170905134 Vol 9 No 5 Oct-Nov 2017 3813