Adaptive Cancellation of Geomagnetic Background Noise Using a Sign-Error Normalized LMS Algorithm

We have presented an adaptive algorithm similar to both the signerror and NLMS adaptive algorithms that is capable of reducing the geomagnetic noise present in magnetic anomaly detection systems. Through both theory and application of this algorithm to actual MAD flight data, we have proven this algorithm’s insensitivity to transients in the desired response, with a corresponding improvement in detection range as shown by receiver operating characteristics. The success of the sign-error NLMS algorithm in this application suggests that this algorithm may also be useful in other situations where impulsive transients are present and must be preserved in the final cancelled output. This work raised a number of issues for the adaptive cancellation of residual noise in MAD systems. In this treatment, a single-reference adaptive filter is used to reduce a single source of noise. However, MAD systems are well-suited to multiplereference adaptive filters, since other auxiliary sensors may be used to cancel the noise related to aircraft motion and equipmentrelated disturbances. Further research shall include the investigation of desired and undesired transients in the reference channel of the adaptive filter and their effects on system performance. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1