The paper presents a framework for autonomous search for a diffusive emittingsource of a tracer (e.g., aerosol, gas) in an environment with an unknown map of randomlyplaced and shaped obstacles. The measurements of the tracer concentration are sporadic,noisy and without directional information. The search domain is discretised and modelledby a finite two-dimensional lattice. The links in the lattice represent the traversable paths foremitted particles and for the searcher. A missing link in the lattice indicates a blocked pathdue to an obstacle. The searcher must simultaneously estimate the source parameters, themap of the search domain and its own location within the map. The solution is formulated inthe sequential Bayesian framework and implemented as a Rao-Blackwellised particle filterwith entropy-reduction motion control. The numerical results demonstrate the concept andits performance.