<![CDATA[Low Rate Wireless Personal Area Network (LR-WPAN) is a communication protocol standardized by the group of IEEE 802.15.4, the focus of its development is to create an architecture that uses low power, has wide range and reliable data transmission. One of its product is Zigbee. Zigbee focus on reliable data transmission could be disrupted if congestion happen due to the density of increased traffic data in its network. Congestion can be prevented by applying the right queue management in its buffer. One of the queue management is weighed random early detection (WRED), the advantage of using WRED is that the mechanism of the queue management not only focus on preventing the congestion but also able to accommodate the type of package that needs to be prioritized in its services. Implementation of WRED in Zigbee was carried out to measure architectural performance for avoiding congestion, which in this study implementation is simulated using Riverbed Modeler simulator. Architecture performance in terms of avoiding congestion was analyzed using parameters of throughput, queue delay, global delay and packet drop. The testing scenario was carried out by comparing WRED to the random early detection (RED) queue management, whereas both queues uses the min thresh and max thresh mechanism to handle congestion. Results of the test concluded that with smaller package size and smaller threshold the use of WRED will be more optimal shown by throughput parameter which is 1.5% better than RED, whereas with larger packet sizes and larger max thresholds the use of RED will be more optimal shown by the throughput parameter values 0.6% better than WRED.]]>
Copyrights © 2019
