| Beidou Navigation System is a new-generation navigation and positioning system all designed by our country.There is far-reaching significance in the study of anti-jamming techniques in it as the electromagnetic environment is more complicated.There are some measures to suppress the in-band interference and jamming such as Least Mean Square (LMS) algorithm which is widely used to suppress the narrow-band interference. As one of classical time-domain adaptive filtering algorithms, LMS algorithm is widely used in digital signal processing designs. However, LMS algorithm is not suitable for high speed processing domains because of its iteration structure. First of all, this paper gives a brief introduction of LMS algorithm. Baseing on relaxed look-ahead technology, it deduces Delayed LMS (DLMS) algorithm and then amend it by introducing delays in both error feedback path and coefficient update loop which can be piplined.Furthermore, DLMS algorithm with limited numerical precision is discussed. According to the expressions of expectation for the coefficient vector and the stationary expectation of the squared estimation error(ESEE) basing on limited numerical precision, it can be concluded that the portion of ESEE deriving from limited numerical precision increases with a decreasing step lengthμor an increasing filter length M. It also introduces leaky factor to lighten the accumulation of bias in coefficient vector.Then, this paper establishes a MATLAB model for the implementation of fixed point DLMS filter in FPGA. The step length, radix point of input signal and coefficient and leaky factor is determined by this simulation model. According to the simulation result, the design of a 17 taps fixed point DLMS filter in FPGA is accomplished. At last, a single channel time domain anti-jamming system is proposed. This paper also expatiates the implementation of the system in FPGA and DSP. After these works, the whole design is completed. |
PDF Full Text Request