Evaluation of the impact of the topology and Hidden Nodes in the performance of a ZigBee Network

Abstract

Low power and small footprint IEEE 802.15.4/ZigBee based devices are a promising alternative to 802.11a/b/g and proprietary protocols for non-critical patient monitoring under important scenarios such as post-op and emergency rooms. However, their use in a healthcare facility to monitor several mobile patients poses several difficulties, mainly because these protocols were primarily designed to operate in low traffic load scenarios. This work presents simulation results used to evaluate the performance of an IEEE 802.15.4/ Zig- Bee based wireless sensors network (WSN) in a vital signs monitoring scenario, for both star and tree based network topologies. The scalability problem in nonbeacon enabled networks is addressed to quantify the degradation in quality of service (QoS) markers when the number of sensor nodes increase. Additionally, the impact of hidden nodes is assessed for the star topology. Results indicate that, to achieve a delivery ratio (DR) higher than 99%, the number of electrocardiogram (ECG) nodes in a star network must not exceed 35. However, considering a tree topology, the maximum number of nodes must be reduced to 18 to maintain the same DR. The network performance is severely impacted by hidden nodes. For instance, in the absence of hidden nodes, a star network consisting of 32 ECG nodes presents a DR higher than 99%; however, if the percentage of hidden nodes is increased to 5%, it drops to 94%. If the same percentage of hidden nodes is maintained, it is necessary to reduce the number of nodes to 13 to reestablish a 99% DR.