Research On Sleep Scheduling For Application In WSN

Due to the special application scenarios, harsh deployed conditions and limited hardwareresources, Wireless Sensor Networks (WSN) face a series of problems of how to reduce thetime of nodes’ idle listening, use the battery energy effectively, reduce the energyconsumption and prolong the lifetime of the network, etc. The more effective and commonsolution is making nodes’ status convert between the sleep state which turning off thecommunication module of the node and the working state based on a certain schedulingperiod. Recently, Chinese and foreign researchers have explored a lot of sleep schedulingtechniques, but all of them have a few disadvantages. This paper discus the sleep schedulingtechniques’ achieving in MAC layer and application layer. In MAC layer, this paper analysisvariety of existing MAC protocols and presents a protocol AD-SMAC which fused Sleepscheduling technology and can dynamically adjust the duty cycle. In application layer, thispaper summarizes a variety of scenarios of WSN applications and presents an adjustableservice-oriented sleep scheduling framework.MAC protocol is the foundation and support research in the field of wireless sensornetworks, its study has a profound influence to the energy management of WSN. This paperstarts to discuss with S-MAC protocol which reflects sleep scheduling firstly, and presentstow research directions based on S-MAC’ two drawbacks. MAC protocol’s development canbe discussed form the synchronous, asynchronous aspect and the duty cycle aspect. This paperanalysis and compares the core mechanism, the performance characteristics and thedeficiencies of the existing technologies, and then presents a protocol AD-SMAC whichconsolidates node’s average packet delay and node’s residual energy.This paper introduces the typical applications of WSN. Based on the summary of theseapplications, this paper extracts the same characteristics and different characteristics tocompare, identifies the key factors which need to be configured for various applications, andproposes two types of sleep scheduling strategy which can meet most applications. Based onthe two types of sleep scheduling strategy, the applications of WSN can be classified into twospecific applications: Class A is a kind of scheduling strategy collecting data periodically.Class B is a kind of scheduling strategy using special event to trigger the alarm. Then thispaper proposes a service-oriented sleep scheduling framework, which designs differentmodule to manage each key factor and leaves interfaces for future factors that may beconsidered. This framework can meet the needs of varies applications by simple configurationand adjust itself according to real-time. It can also avoid wasteful duplication, improve workefficiency and provide a good solution to the issue of application independent of sleep scheduling in WSN.The simulation experiment and data analysis are made for two typical applicationscenarios in the network delay and energy consumption aspects. The experimental resultsindicate that, no one MAC protocol can provide a perfect performance for all of the WSN’sapplication scenarios. The simulation experiment and data analysis are made for S-MACprotocol with different set of duty cycle. The experimental results indicate that, even the sameprotocol, as long as the key factors are different, the application scenario which the protocol issuitable for is different. The simulation experiment and data analysis are made for comparingAD-SMAC with S-MAC and DS-MAC in terms of network latency and energy consumption;the experimental results indicate that AD-SMAC has a better performance compared toS-MAC and AD-MAC. This paper proposes the work module, designs the frame format forthe framework, and then gives a way to realize the framework using TI CC2430or CC2431chipsets with Z-stack. The experimental results indicate that the sleep scheduling frameworkrealized in this paper is able to configure the sampling interval, alarm threshold convenientlyand the sleep scheduling framework successfully realizes the dynamical adjustment ofsampling interval, alarm threshold.