To solve the problems of network management and security monitoring and auditing in trusted network, by researching the framework of trusted network connect (TNC) and the architecture of cyberspace situational awareness (CSA), in view of the certain and uncertain characteristics of multi-source information in trusted network security, a new security situational awareness and forecast based on set pair analysis (SPA) called SPSAF is proposed. SPSAF audited network connection information, system management information, system monitoring information and application service information by the method based on feature base. The security situational indexes were extracted by the method of improved entropy method (EM) and analytic hierarchy process (AHP). The SPA method was utilized to assess these indexes to get the value of network security situation, and the security-situation-graph of network was drawn. The time series of the assessment results was analyzed via the Box-Jenkin model to forecast the network security trend. The simulation experiment results show that SPSAF is accurate and effective to reflect the network security situation and its trend, which can help the administrator to make security policies efficiently, to discover risks timely, to adjust policies and take relevant protective or emergency measures quickly and accurately, and to ensure overall and reliable security.