The protection mechanisms of the SMIMS against different types of attacks are summarized in Table 3, and the details of analyses for each type of attack are illustrated below.
The computational complexity of the SMIMS is analyzed to demonstrate its efficiency.
In the SMIMS, our modules are implemented in the on-body sensors and the medical databases, without incurring any additional hardware cost.
In the following sections, the performances of the SMIMS are tested, including the security issue, memory consumption and computation cost.
In this section, the security performance of the SMIMS is tested.
Therefore, the confidentialities of the biometric data and EMR are significantly protected in the SMIMS.
1 and indicate that the threats of data modification and impersonation attack can be detected and prevented in the SMIMS.
In this section, the memory consumptions of SMIMS are examined and compared with those of the system without security or medical management function whilst the lengths of biometric data are set to different values.
It can be observed from Figs 5-6 that the RAM consumption rates of our SMIMS are less than 50%, which is 2048 bytes, and the ROM consuming rates are even lower.
It is demonstrated that the memory consumptions of all the security and medical management processes are quite low and that the memory consumption of the whole SMIMS is acceptable based on the above results.
The computation cost of the SMIMS includes 3 parts: the execution time of the EMR prewriting, the EMR sharing and the EMR identification processes.
3 and indicates that the proposed SMIMS is of high-efficiency.