نوع مقاله : علمی-پژوهشی
نویسندگان
1 دانشکده مهندسی کامپیوتر دانشگاه اصفهان
2 دانشیار گروه مهندسی فناوری اطلاعات - دانشگاه اصفهان
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
One of the services that 5G provides is URLLC. To meet the requirements of URLLC, Grant-Free access schemes have been proposed. In these schemes, the radio resources are utilized by User Equipments without any reservation . In these scenarios, we may see malicious behaviors such as the inclination of UEs to reduce their own latency with selfish behaviors or possibility of having misbehaving nodes that degrade the QoS of others. Therefore, detecting malicious users is important in these schemes. In this research, we seek to detect misbehaviors in K-repetition based slotted-ALOHA grant-free access scheme. In previous studies, the security of traditional MAC protocols such as ALOHA has been studied. However, the security of grant-free access based on ALOHA mechanism has not been investigated. Deep learning is considered for misbehavior detection and we employ a Long Short-Term Memory (LSTM) model to detect malicious UEs. The simulation results indicate that under stable conditions, the accuracy decreases as the number of nodes increases. Additionally, in the presence of a fixed number of nodes, as the number of malicious nodes increases, the accuracy decreases. Additionally, if malicious nodes behave properly at times and misbehave at other times, detecting such misbehavior is more difficult.
کلیدواژهها [English]
)