RFID tag oriented data allocation method using artificial immune network

Radio frequency identification (RFID) enables a seamless link between the patient data stored on RFID tag and the medical monitor, which provides an instant access to the relevant information for healthcare services. With the patient tag, incorrect data inputs can be prevented and errors in patient treatment can be detected in real-time. However, only important data items can be allocated to a RFID tag with the limited memory. In general, the RFID tag oriented data allocation problem can be mitigated by minimizing the total value of “unexplained” data off tag (TVUD) which is related to the memory capacity and the correlation matrix. Artificial immune network is an emerging heuristic algorithm that is broadly used to solve scientific researches and engineering problems. This paper formulates the RFID tag oriented data allocation problem as a nonlinear knapsack problem and proposes an artificial immune network (DA-aiNet) to solve this optimization problem. A series of numerical experiments are arranged to investigate the effects of memory capacity and correlation matrix. Further experiments are used to make some comparisons between the proposed DA-aiNet and the other existing algorithms. The experimental results indicate that this proposed DA-aiNet is more efficient in minimizing TVUD than the particle swarm optimization and the genetic algorithm.