The future war represented by high-precision guided weapons and unmanned systems has a high demand for satellite navigation signals. Based on the theory of satellite navigation spoofing interference， the concept of agile navigation warfare supported by low-orbit constellation is proposed. The definition and connotation of the concept are given， and its winning mechanism is further explained. By installing spoofing interference equipment， signal suppression and spoofing with more maneuverability， flexibility， and wider coverage can be achieved. Based on the concept of agile navigation warfare， specific combat scenarios are designed from the aspects of combat objectives， battlefield environment， combat forces， combat operations， combat effects， etc. With multi-means support in the sea， land， air and other combat domains， through adaptive control of spoofing signal power and dynamic adjustment of spoofing strategies， it ensures continuous and stable spoofing interference signal output， achieving regional signal suppression of blue threat targets and gradual covert inductive spoofing and deviation. This provides capability to support the red side to obtain the battlefield initiative and effectively carry out defense and attack missions. The key technologies that need tobe broken through in the implementation of agile navigation warfare and their feasibility are analyzed and explained.

Ultra-long-range air defense is committed to preventing the threat target from thousands of miles away， and requires close cooperation between the three chains of strike chain， accusation chain and intelligence chain. Aiming at the ultra-long range air defense， an innovative warfare method that changes the rules of war， this paper first analyzes the various processes of strike chain， accusation chain and intelligence chain， and then puts forward the space-based information direct support mode. The application of space-based information direct support in ultra-long-range air defense operations is studied， and how space-based information direct support enables strike chain， accusation chain and intelligence chain is deeply discussed， in order to provide reference for the construction and development of ultra-long-range air defense combat information support systems in the future.

Aiming at the lack of automated and intelligent forensics methods in the aviation reconnaissance and evidence collection， an aviation forensics target detection algorithm based on the fusion of RetinaNet and SE is proposed to solve the problem of large changes in target scale and imbalance in data collection. The performance is further improved by introducing an attention mechanism through the SE （squeeze-excitations） module. The feature pyramid network（FPN） can effectively deal with the problem of large changes in target scale， Focal Loss can effectively deal with the imbalance of categories in dataset， the SE module introduces a channel attention mechanism to strengthen the feature map， and can further use the extracted channel correlation to enhance effective features and suppress ineffective features. Through simulation experiments， it is verified that the algorithm can improve the accuracy of target detection with a small amount of calculation increased， further enhance the characterization ability of the model， and effectively improve the efficiency of target detection， which can provide a reference for related engineering applications.