Research On The Algorithms For Collecting Sensory Data

Sensory data refers to the physical world real-time running status data collected by sensory nodes such as sensors or sensing devices,and sensory data are generally collected by a network which consists of a large number of sensory nodes(“Sensory Network”for short).Thanks to the rapid development and gradual maturity of communication,sensor and microelectronic technologies,the micro sensors which have the ability of computing,sensoring,and communication are developed.These sensors include smartphones,tablets and other intelligent terminals.If a large number of traditional micro sensor nodes are deployed on stationary objects such as forests and buildings,a static wireless sensor network will be formed.If a large number of sensors are placed on moving objects,a mobile wireless sensor network will be formed.At present,sensor networks have been widely applied in the fields of environmental monitoring,target tracking,traffic supervision,defense military,health care,and disaster prevention.In addition,the smartphones integrated with multiple sensor devices form the mobile sensing network,which can sense to the activity status of the person and activity area.In the absence of cellular network or no signal,the mobile sensing network based on bluetooth or WI-FI Direct can be used for data sensing.Collecting sensory data is the basis on event detection,event prediction,event mining and decision analysis,and one of the most basic and the most important research problems in wireless sensor network and smartphone mobile sensing networks.Collecting sensory data is the main research topic in this paper.The main contributions of the paper are given as follows:Firstly,the problem of sensory data collection is studied based on multi –power and multi-channel in static sensor networks.By using BloomFilter data structure to store data,the nodes can easily merge the data transferred from other nodes without additional storage when the data are collected.A distributed data collection algorithm with dual-power and multi-channel is presented.The algorithm distributedly builds a cluster and the cluster head is responsible for calculating data merge scheduling in each cluster.Data merge can be performed in parallel among clusters and the collector calculates the data merge scheduling among clusters and then collects the data of each cluster head.This paper theoretically proves and gives a delay limitation of data collection.The simulation results show that the average data collection latency of the algorithm proposed is smaller than that of single channel and multi-channel scheduling algorithm with minimum delay in theoretical analysis.Then,the problem of sensory data collection is studied based on probability guarantee in the mobile sensor network.For the high-speed mobile sensor network,a data collection algorithm named DCPG based on probabilistic guarantees is proposed.The theoretical analysis shows that DCPG can adjust the system parameters to guarantee that the average data collection ratio is greater than the user-specified data collection ratio ?.DCPG can guarantee to have a higher average data collection rate in dense high-speed networks,while significantly limiting communication overhead.The simulation results further show that DCPG outperforms the existing best data collection protocol SMITE on the aspects of data collection ratio and communication overhead for dense networks.DCPG algorithm is suitable for high-speed and dense opportunities networks.Next,this paper studies the feasibility of collecting sensory data in smartphone mobile sensor networks based on bluetooth.This paper makes full use of the feature of bluetooth matching connection in smart phone,and uses BloomFilter structure in bluetooth name to optimize connection efficiency.Based on active connecting pattern,the algorithm ACDC is proposed.Also,algorithm HDCA based on active and passive combined pattern is presented.Experimental results show that HDCA can avoid invalid data transmission,reducing redundant traffic,collect all data using the lowest number of connections.Finally,this paper studies the feasibility of collecting sensory data in smartphone mobile sensor networks based on WI-FI Direct.Two data collection algorithms based on proactive request and passive acception are proposed.Also,the influence of seed nodes on data collection rate is analyzed.If some active nodes are selected as the seed nodes,the rate of data collection can reach more than 90%,so the selection of the seed nodes is crucial for data collection.When using the same number of seed nodes,the proactive request algorithm is better than the passive acception in data collection rate.