In order to solve the location safety problem of fire rescue firefighters,a collaborative location algorithm of forest fire based on long range radio(LoRa)technology was proposed.Firstly,aiming at the non-line of sight (NLOS) problem in field rescue,an extended Kalman filter cooperative location algorithm with modified covariance matrix was proposed,assisted by the ranging information between LoRa mobile nodes in the cooperative network.The system equation was established and the error covariance was updated twice according to the current state filtering value to correct the deviation.Secondly,an improved median filtering algorithm was proposed,which has used dynamic window screening to process the location disturbance data and optimized the location trajectory curve twice.Simulation results indicate that the proposed algorithm greatly coordinated the entire network resources,and the positioning accuracy can reach 0.49 m in the outdoor environment with multiple NLOS and poor communication quality,which has provided a reliable solution for forest fire safety rescue.