Sun Yat-senuniversity

Xingcheng Liu

Yitong Liu

Qiang Tu

Yi Xie

Shuo Liang

Zhao Tang

Yingying Zhao

Shizhan Cheng

Hao Lin

A method of hop count matrix recovery based on a decision tree classifier, includes: S: performing a flooding process to acquire a hop count matrix {tilde over (H)} with missing entries; S: constructing a training sample set according to relationships between a part of observed hop counts in the hop count matrix {tilde over (H)}, and modeling the observed hop counts in the hop count matrix as labels of the training sample set, wherein a maximum hop count represents a number of classes; S: training a decision tree classifier according to the training sample set obtained in step S; and S: constructing a feature for an unobserved hop count, to obtain an unknown sample; and inputting the unknown sample to the trained decision tree classifier, to obtain a class of the unknown sample which represents a missing hop count at a corresponding position in the matrix, to recover a complete hop count matrix H.

Method of hop count matrix recovery based on decision tree classifier

In a data processing method for node localization in wireless sensor networks, a plurality of potential coordinate calculation values for a to-be-located node are calculated based on RSSI data received by the to-be-located node from a plurality of anchor nodes. A centroid Cof these coordinate calculation values is calculated. Then coordinate calculation values with distances to the centroid Cless than a threshold are screened out, and a centroid Cof the screened-out coordinate calculation values is calculated. A coordinate position of the centroid Cis used as the position of the to-be-located node. In other words, a plurality of potential positions for the to-be-located node are calculated using multiple groups of RSSI data and these potential positions are then screened through calculation. Instead of directly processing the received RSSI data, the present disclosure processes the coordinate calculation values obtained based on the received RSSI data.

Data processing method for node localization in wireless sensor networks

A wireless sensor network (WSN) clustering routing method based on optimization with salp swarm algorithm includes: S1: constructing an objective function by comprehensively considering factors as follows: node residual energy, a distance between cluster head nodes, a distance between a cluster member node and a cluster head node, a distance from a node to a base station, and a number of cluster head nodes; S2: solving the objective function by using a salp swarm algorithm, to find an optimal clustering scheme, and thus dividing the WSN into non-uniform clusters, wherein in each cluster, the cluster member node sends data to the cluster head node in a single-hop manner; S3: constructing a minimum spanning tree with the base station as a root node by using a routing utility function &#x192;(CH); S4: sending, by the cluster head node, data to the base station through routing organized by the minimum spanning tree.

WSN clustering routing method based on optimization with salp swarm algorithm

A malicious anchor node detection and target node localization method based on recovery of sparse terms, includes: S1: establishing an unknown disturbance term by using ranging value attack terms from an attacker to nodes in a wireless sensor network, and introducing a to-be-estimated location of a target node to the unknown disturbance term, to obtain an unknown sparse vector; S2: converting a problem of malicious anchor node detection and target node localization into a problem of recovery of the unknown sparse vector; S3: determining a location of an initial node according to a recursive weighted linear least square method, and recovering and reconstructing the unknown sparse vector with sparsity; and S4: determining a malicious anchor node determination range by approximating a threshold using a recovered value of the unknown sparse vector, to implement malicious anchor node detection, and recovering and determining location information of the target node.

Malicious anchor node detection and target node localization method based on recovery of sparse terms

A decoding method of low-density parity-check (LDPC) codes based on partial average residual belief propagation includes the following steps: S1: calculating a size of a cluster &#x3c0; in a protograph based on a code length m and a code rate of a target codeword; S2: pre-computing an edge residual rcorresponding to each edge from a variable node to a check node in a check matrix H; S3: calculating, based on &#x3c0;, a partial average residual (PAR) value corresponding to each cluster in the check matrix H; S4: sorting m/&#x3c0; clusters in descending order of corresponding PAR values, and updating an edge with a largest edge residual in each cluster; S5: updating edge information mfrom a check node cto a variable node v, and then updating a log-likelihood ratio (LLR) value L(v) of the variable node v; and S6: after the updating, making a decoding decision.

Growing community of inventors

Guangzhou, China

Xingcheng Liu