HFSS results are more accurate than CST because HFSS is a FEM software, whereas CST is a FIT. I have validated this fact on a number of geometries, but for your case CST should also be fairly accurate
I have designed single-layer microstrip antennas, Proximity-coupled microstrip antennas (with slits and slots also) and Aperture-coupled microstrip antennas of various types using IE3D. I found that IE3D results are agreeing well with measured results.
You can use either of them but in most cases CST is found to be the most favorable one for micro strip patch antennas since it uses Finite Integral Method (FIM) and use less simulation time to bring out the simulated results.
CST MWS is time domain analysis tool i.e.. transient solver capable of providing broad frequency domain result like S parameter for large range of EM problems. Furthermore its more efficient in term of memory utilization and time for solving any particular problem i.e planar antennas
I and many co-authors have being using CST microwave studio for the last 7 years. The obtained results compared well with experimental results. In many cases we found that CST and HFSS gave similar results. For better and fair comparison, one should use similar accuracy limits.
Both softwares have their own computational methodologies so we can't say one is better than the other and both are used by a large community of researchers.
However for those new to this area, CST interface is much more convenient to design simple micro-strip antennas, its template based design procedure automatically includes the boundary conditions and results are available in different categories, the desired ones can easily be observed.
HFSS on the other hand provides more design options with it design interface but require some understanding of using appropriate boundary conditions and performing simulation.
I have used CST software in my PhD thesis and in my papers, I found that CST have more flexibility, simplicity, and speed is higher than HFSS which I used it to validate CST results. on the other hand, the boundary condition, port excitation, and interfacing wiith MATLAB are easy to use than HFSS. Finally, CST is better than HFSS in terms of speed for large iterations of sweeping and optimization process.
CST and HFSS are the better softwaire, I think for a simple microstrip patch you can do it with CST it was faster in the design (you don't need to plot the port, dimensions are fixed unlike HFSS)
The part of the user guide is patch antenna design if you are not able to find I am able to send a design manual or the complete sample of patch antenna
i think it depends on the nature of your design ,as we know each software uses a specific numerical method for analysis and simulation , so based on the most theoritical numerical method that could give you best theoritical answer choose your simulation software .that's how a good designer think .ex : for more geomitric details its best to either choose hfss or CST ,for large reflector antennas best to use FEKO or IE3D
The question should be "which software is more user-friendly and can be learned more easily and quickly". Because most of the soft wares mentioned are capable of antenna design and optimization. I really like Ansys HFSS because of its antenna design toolkit.
You know, most of software developing companies are pushing hard to get more stable and efficient numerical solvers and make them more and more user friendly. they also say many publications cite them because researchers and companies use their soft wares. You should check their sites.