Minimising collisions in RFID data streams using probabilistic Cluster-Based Technique

Radio Frequency Identification (RFID) uses wireless radio frequency technology to automatically identify tagged objects. Despite the extensive development of the RFID technology in many areas, tags collisions still remain a major problem. This issue is known as the collision problem and can be solved by using anti-collision techniques. Current probabilistic anticollision approaches suffer from tag starvation due to the inaccurate Backlog estimation and have a low performance in some cases. In this research, we propose a Probabilistic Cluster-Based Technique (PCT) to maximise the performance efficiency during the tag identification process. The PCT approach creates new tag grouping strategies using particular equations, according to the optimal efficiency obtained for a specific number of tags. Through extensive experimentation, we have demonstrated that the proposed concept performs better than the other current state-of-the-art approaches.