Research On The Algorithms In Adaptive Interference Canceller

In modern military communication, due to the presence of multiple devices in a small space, the transmitter and receiver antennas are often relatively close. In addition, the receiver and the transmitter have a large difference in level, so the transmitter produces a strong electromagnetic interference to the receiver. This kind of interference is called cosite interference. It causes the receiver desensitization or even blocking, affecting the performance of receiver seriously. There are many ways to solve this problem. The most effective way of them is adaptive interference canceller. In addition, this system can be assembled in electronic jamming aircraft to keep the communication capability, while jamming enemy communications. At last, this system can also be used to implement duplex communication in co-frequency and co-time. At present the interference signal is sharply cancelled mainly through the combination of antennas layout design and interference cancellation in the analog and digital domain. This paper studies interference cancellation in the analog domain.This paper first describes the research background and the applications of the adaptive interference canceller. At present, this technology is mainly used in military domain, but its development in civilian use is very promising. Then this paper introduces the principle of adaptive interference cancellation. The adaptive interference canceller is a specific application of adaptive filter. It synthesizes a signal having equal amplitude and opposite phase with interference signal to cancel the interference itself. According to the effects of amplitude deviation and phase deviation to interference cancellation ratio, a kind of adaptive interference canceller based on orthogonal vector synthesis is designed and its modules are introduced. Then a scientific analysis is conducted of data extracted from the control unit. Simulated annealing algorithm, genetic algorithm and pattern search algorithm are proposed as the control algorithm. The pattern search algorithm is improved. Then data extracted from the actual system is used to simulate and analyze above algorithms. The results show that, compared with simulated annealing algorithm and genetic algorithm, pattern search algorithm has faster convergence speed, while convergence accuracy is almost the same. After improved, pattern search algorithm's convergence speed and convergence accuracy are strengthened. Finally, the improved pattern search algorithm is implemented in the actual system and the interference cancellation ratio is measured in 30~512MHz. The results show that the ICR can reach more than 40 d B in all frequencies, which can satisfy the demand of interference cancellation.It can be seen, adaptive interference canceller using the improved pattern search algorithm can solve the problem of cosite interference. In addition, pattern search algorithm has the advantages of strong local search capability, fast convergence speed, easy implementation in project. After improved, the algorithm's convergence speed is further accelerated and the global search capability is strengthened. This algorithm is suitable for solving optimization problems in project requiring fast convergence.