Community network management (RT RW Net) faces scalability challenges in bandwidth management as the number of hotspot voucher users increases. Traditional methods using static limitations on MikroTik RouterOS are often inefficient and burden router CPU performance. This study proposes a Quality of Service (QoS) optimization solution through the integration of dynamic RADIUS attributes with the Per Connection Queue (PCQ) algorithm. Using the Network Development Life Cycle (NDLC) methodology, this research implements the delivery of Vendor Specific Attributes (VSA) from the RADIUS server to automatically classify traffic into PCQ queues based on a network topology consisting of one gateway router and several access points. The use of MikroTik User Manager's internal feature as a local RADIUS server facilitates centralized voucher management. Test results show that this method can increase bandwidth utilization efficiency by up to 26% under peak load conditions, maintaining jitter values below 5 ms, average delay of 42 ms, and packet loss of only 0.2%. Furthermore, the use of dynamic attributes reduces administrative complexity as the system automatically provisions queues without manual operator intervention.
Copyrights © 2026