This paper presents an advanced current sensorless control strategy for the individual Maximum Power Point Tracking (MPPT) on a Cascaded H-Bridge Multilevel Inverter (CHBMLI) with 21-level (10 H-Bridges). The CHBMLI is controlled by a real-time distributed architecture where each H-Bridge is connected to a single solar panel and is controlled by a local controller. All local controllers communicate with a Master controller through a real-time robust fieldbus. The control scheme, presented in this paper, regulates each solar panel to its own Maximum Power Point (MPP) thus maximizing the harvested energy even with mismatch conditions in the photovoltaic (PV) array. It improves the working of the well-established Perturb & Observe (P&O) method by reducing power oscillations. This increases the MPPT efficiency η MPPT. Furthermore, the local controller estimates the local current of the solar panel by computing a state observer based on physical equations and by taking advantages of locally accessible data. The benefit of this state observer is to suppress the requirement for local current sensors. This leads to reduced costs and simplified local controllers. The proposed method is validated with MATLAB simulation followed by an experimental setup on a 130 V rms grid voltage with 6 individual modules. The experimental results demonstrate that the complete control strategy can be implemented on low cost Micro Controller Unit (MCU).