Performance Evaluation of Queue Algorithms for Video-on-Demand Application

Nowadays, there are real-time internet applications that require low latency and high bandwidth usage. In order to satisfy this requirement, many active queuing algorithms are offered, such as RED, BLUE, and controlled delay (CoDel). The number of publications aimed at evaluating the performance of the active queue management (AQM) algorithm is still limited by using a network simulator and only comparing two algorithms. The combination of AQM and connection types (fairness and no-fairness) in real network implementation has not been discussed much. In this paper, we offer a comparison of the performance of three queuing algorithms, namely DropTail, BLUE, and CoDel, with two types of connections, namely fairness and no-fairness using real network devices and internet video streaming access. Experiments carried out in the real network environment using Sandvine PL8840's bandwidth management device. The quality of services (QoS) parameters that will be compared are throughput, latency, and packet loss. The comparison results show that fairness connections on all queuing algorithms can maintain low latency values; throughput is smaller than no-fairness connections. The absence of packet loss in streaming video based on the TCP protocol does not guarantee that the video being played runs well because it depends on the available bandwidth capacity.