Performance Evaluation on IEEE 802.15.4 Scatternet Sensor Networks

Jun Fung · Mspace (University of Manitoba) · 2007

Low Rate-WPAN networks are equipped with battery-powered sensors and chang- ing batteries is considered impractical.Therefore, energy saving is very important when building such networks.IBEE 802.15.4 is a standard for LR-WPAN, which specifies the Physical layer and MAC layer for LR-WPAN.Scatternet is a network structure where sensors are grouped into clusters using star topologypiconets.Each piconet has a coordinator as the communication center.Two piconets are connected by an intermediate node -bridge.In my thesis work, I have investigated the feasibility of building a large-scale LR-WPAN scatternet network using a simulation model.My work can be divided into three phases.In the first phase, I studied the performance of the bridge communication by simulating a two piconets model where piconets are interconnected by a Master/Slave bridge.I evaluated the performance of the inter-cluster communication and found the performance saturation mode is reached when the traffic load is high enough in each cluster.I also compared the performance of bridges using different communication protocols and parameters andfound that in different traffic load we can choose diferent potocols to achieve best performance.In the second phase, i applied a duty cycle technique to each node to Abstract let the network achieve the prescribed event sensing reliability by a distrìbuted activity management.In order to balance the lifetime of all the clusters in the network, a nocie compensation method was proposed to extend the lifetime of heavily loaded clusters.In the third phase, I proposed a scalable, energy-efficient and lifetime-baiance routing algorithm for the network to extend the network lifetime as much as possible.My simulation results show that when the network is using this routing algorithm, the workload of clusters are balanced and the network lifetime is extended effectively.

Read the paper · More papers on PaperTik