The Data Collection Scheme Towards Opportunistic Connections In Wireless Sensor Networks

Wireless sensor networks(WSNs)are composed of various heterogeneous sensor nodes,and they are used in life,industry,and military widely because of their advantages of large-scale deployment and self-organizing networks.Due to limited resources and different capabilities,sensor nodes could perform different data sensing tasks,therefore,how to collect data from sensor nodes efficiently is a hot issue worthy of studying.In the practical application of WSNs,it is difficult to ensure that the neighboring nodes are 100% connected,and the reasons for the connectivity between adjacent nodes are various,which can be summarized as follows:(1)WSNs are usually deployed in harsh environments,affected by which,the unreliable links could exist between any two adjacent nodes;(2)The movement of nodes usually leads to intermittently connected links and creates some forms of opportunistic connection;(3)Sensor nodes are usually powered by batteries,to save their limited energy,the nodes could adopt sleep mechanisms,in which the adjacent nodes can communicate with each other only when both of them are all in working status;(4)With the intellectualization of WSNs,sensor nodes could have some new features,such as social attributes,it will bring opportunistic connections between adjacent nodes.All of the above could lead to opportunistic connections between adjacent nodes,which further brings a challenge in the data collection.To cope with this challenge,the paper first introduces random graph theory to model WSNs to realize the analysis of network connectivity,in which the Opportunistic Connection Random Graph(OCRG)is constructed based on that nodes adopt the asynchronous working-sleeping cycle strategy,and the Smart Node Random Graph(SNRG)is constructed based on the link quality of adjacent nodes and the social attributes(receiveing or forwarding packets according to their preferences)of a node.Then,based on SNRG,the paper further studies the connectivity optimization and realizes the connectivity determinition of any source-destination nodes with the minimum test cost,in which the connectivity of any adjacent nodes is assumed to be tested and the test cost is the same.Next,the paper proposes two data collection schemes,the first one considers that the unreliable links of adjacent nodes result in the opportunistic connections between tham,based on which,the data collection towards links unreliability is proposed.In the scheme,the mobile sink with a random track is adopted to collect data from sensor nodes,during the date delivering from the nodes to the sink,the data forwarding capability of nodes is calculated to select the forwarder,which improves the efficiency of the data collection.The second one considers that the node adopts the asynchronous working-sleeping cycle strategy results in the opportunistic connections between adjacent nodes,based on which,the data collection towards nodes sleeping is proposed.The scheme also adopts the mobile sink with a random track for data collection,to deal with the challenge brought by the node status transformation,OCRG is constructed to calculate the data forwarding path for the nodes in the data collection scope in advance,thereby improving the data transmission success ratio and reducing the energy consumption.The experimental results show that the proposed connectivity determination method is effective,and the two data collection schemes have good performance in packet delivery ratio and energy consumption.