This thesis is focused on autonomous trajectory planning for Active Debris Removal (ADR), commencing from far-range rendezvous and up to a point just before docking. Unlike typical rendezvous missions, the target in ADR is not designed for rendezvous and capture; it is usually uncooperative and possibly tumbling. If the satellite is unpassivated, then there is the added risk of explosion, while the presence of large appendages can make the satellite fragile. A case in point is Envisat, European Space Agency's largest satellite (about 8000 kg) that went defunct in 2012 and is currently, their prime concern for ADR. Most past and current technology for guidance is based on the Clohessy-Wiltshire (CW) equations, which are limited in terms of application - they use linearized dynamics, need circular orbits, employ impulsive burns, and require the user to provide some initial guess for the solution. Realistic final approach involves a number of trajectory constraints, like those on docking axis alignment, approach cone angle, keep-out-sphere, station-keeping and hold points, which are hard to include with CW-equations. Further, there is a need to make the system autonomous so that it can handle off-nominal situations as well. Convex optimization problems are a class of problems that are nearly as easy to solve as linear problems, but are more general in their scope of application. They can handle some non-linear constraints, be solved in polynomial time and are guaranteed to give global minimum. Inspired from the recent development of convex-optimization based trajectory-planning technique by Liu (Autonomous trajectory planning by convex optimization, PhD thesis, Iowa State University, 2013), a guidance algorithm is developed for the Envisat mission. In comparision with Liu's algorithm, it shows faster convergence (upto 5 times faster), improved accuracy (by an order of 2), and low computational cost (upto 40% in specific cases). Additionally, a method to automate the problem formulation in conic form is provided, so interfacing software can be skipped and the problem can be solved upto 9 times quicker. The guidance algorithm is able to handle all constraints of the Envisat mission, with the exception of forced motion constraint inside the keep-out-sphere. An alternative formulation is suggested which is moderately successful. The algorithm is also able to generate solutions for various attitude scenarios of Envisat. However, it is found that the method is not reliable, as some times feasible problems are found to be infeasible, or the solution is over/under estimated. Also, despite the improvement in accuracy, it still deviates substantially from the computed path, and needs a path-tracking mechanism. A PID controller is found to be sufficient for this purpose. A novel work-around to include perturbations was successfully demonstrated as well. To summarize, although the convex-optimization based guidance method has very specific advantages over CW-equations, there is a lot left to be improved with respect to robustness and autonomy.