This Thesis considers the use of robotic swarms with limited exteroceptive and inter-robot sensing capabilities for exploring and mapping unknown environments. It presents a novel collaborative exploration strategy that directs the resource-constrained swarm toward unexplored areas in the environment by dividing the swarm into two teams:​ landmark robots and mapper robots, respectively. The former directs the latter toward ‘promising’ frontiers in the environment enabling efficient frontier-based exploration. The landmark robots are optimally positioned to maximize new information added to the map while also adhering to connectivity constraints. The mapper robots are decentralized and use random motion to explore the local area surrounding the landmark robots. Through this decoupling of the swarm into two separate teams, directing the swarm to new areas of the environment can be addressed separately from exploring and mapping that new area. The performance of the exploration and mapping strategy has been validated through extensive simulations.