The Cluster mission allows the determination not only of the bow shock crossing position but also, with a simple timing method and a reasonable confidence, the shock normal and the velocity along this normal. We apply this technique to a series of eleven consecutive bow shock crossings which occurred during a time interval of approximatively two and a half hours on 31 March 2001. We fit, on a distance versus time frame, the position of the bow shock subsolar point by imposing that the time derivatives at the crossings be equal to the shock speeds we determine. The curve we obtain this way represents global oscillations of the bow shock with a typical amplitude that compares quite well to the prediction of standard gasdynamic models which take into account the upstream solar wind plasma conditions.