Focusing on the issue of accurate assessment of the communication nodes in Wireless Sensor Network (WSN), a Grey Theory-based Trust Model named GTTM was proposed. The proposed model fully monitored the behaviors of nodes in network, builded sample matrix, and computed the weights of the recommending nodes by grey relational method, and computed nodes' trust values through the clustering algorithm of grey theory. The simulation experiments showed that compared with the classic Reputation-based Framework for Sensor Networks (RFSN) model, the convergence of the trust value of nodes in GTTM network is more gentle; GTTM could resist malicious recommendation and reduce the values of untrusted nodes in a timely manner, could still get a high rate of successful trading when the network suffered attacks. Compared with the Trust Computation Model based on Bayes Estimation (TCM-BE) model based on Bayes estimation, even under the circumstances of less recommended samples, GTTM still could keep low rate of false positive malicious nodes. The experimental results show that the GTTM can evaluate the trust values of nodes more accurately and ensure the reliable operation of the network.

The static task list scheduling problems in distributed heterogeneous computing environment of Internet of things was studied, and a list scheduling algorithm named Heterogeneous Dynamic Priority Task Scheduling (HDPTS) was proposed, which can dynamically change scheduling sequence based on the strategy of the earliest completion time. Concerning that the exsiting list scheduling algorithms can not accurately determine the scheduling order before scheduling, on the basis of Improved Heterogeneous Earliest Finish Time (IHEFT) algorithm, a dynamic priority scheduling policy was added to it. When precursor tasks of a node completed scheduling, the scheduling priority of this node should be changed. Scheduling priority of task was calculated on the basis of choosing the maximum value between the latest completion time of all immediate predecessor tasks and the maximum available time of all the resources. At the same time, some other factors were also considered, including the influence to the subsequent tasks of the tasks assigned to the resource, the resource load, the calculated value of uplink weight and the influence to the exit tasks. All these considerations make the priority calculation be more reasonable, so as to dynamically change the task scheduling sequence reasonably according to the task allocation situation. By a randomly generated example test, the results show that the scheduling length of HDPTS reduced by 14.29% compared with IHEFT, HEFT (Heterogeneous Earliest Finish Time); the test results on a large number of randomly generated Directed Acyclic Graph (DAG) with specific structure prove that HDPTS is more effective than IHEFT, HEFT and LDCP (Longest Dynamic Critic Path) algorithms.