The AWAKE experiment relies on the self-modulation instability of a long proton bunch to effectively drive wakefields and accelerate an electron bunch to GeV-level energies. During the first experimental run (2016-2018) the instability was made phase reproducible by means of a seeding process: a short laser pulse co-propagates within the proton bunch in a rubidium vapor. Thus, the fast creation of plasma and the onset of beam-plasma interaction within the bunch drives seed wakefields. However, this seeding method leaves the front of the bunch not modulated. The bunch front could self-modulate in a second, preformed plasma and drive wakefields that would interfere with those driven by the (already self-modulated) back of the bunch and with the acceleration process. We present studies of the seeded the self-modulation (SSM) of a long proton bunch using a short electron bunch. The short seed bunch is placed ahead of the proton bunch leading to self-modulation of the entire bunch. Numerical simulations show that this method have other advantages when compared to the ionization front method. We discuss the requirements for the electron bunch parameters (charge, emittance, transverse size at the focal point, length), to effectively seed the self-modulation process. We also present preliminary experimental studies on the electron bunch seed wakefields generation.