In the multimodal biometric system, the parallel fusion mode has more advantages than the serial fusion mode in convenience and efficiency. Based on current works on serial multimodal biometric system, a framework combined with parallel fusion mode and serial fusion mode was proposed. In the framework, the weighted score level fusion algorithm using biological features of gait, face and finger was proposed at first;then semi-supervised learning techniques were used to improve the performance of weak traits in the system, and the simultaneous upgrade of user convenience and recognition accuracy was achieved. Analysis and experimental result indicate that the performance of the weak classifier can be improved by online learning, the convenience and recognition accuracy are successfully promoted in this framework.

To prevent eavesdropping in Orthogonal Frequency Division Multiplexing Vehicular Ad-Hoc Network (OFDM-VANET), a cross-layer security routing protocol was presented in this paper, which includes two parts: secure routing protocol and secure frequency diversity. On the one hand, the eavesdropping of the unrelated node was prevented, by limiting its knowledge on information transmission routing. On the other hand, in the diversity of physical layer receiving, the fading difference among wireless channels, which was transparently transmitted from network layer, was utilized to identify the destination node and eavesdropping nodes. The theory analysis and simulation results illustrate that, when the vehicle on a scale of 1000 to 2000, the protocol makes instantaneous eavesdroppers only about 3. Compared to the conventional methods, it can effectively reduce the number of eavesdroppers.

The existing sequential acquisition methods search the correlation peak continuously in receiving signals, thus wasting a great deal of calculation resources to search correlation peak in the received signal containing no target sequence. In another case, with the sequence period increasing, the calculation resources perform exponential growth. As a result, the target sequence can be hardly got. In order to reduce the calculation resources, the paper presented a sequence core transform method to acquire the sequence. The method compressed the entire sequence information into the local sequence core, and by this sequence core, the acquiring process was divided into two steps: sequence detection and sequence acquisition. In this process, the calculation resources were reduced because the sequence acquisition only occured in the second step. Furthermore, by using the down-sampling on received signals, the sequence core transform could decrease the sequence amount needed in the acquiring process. The theoretical analysis and simulation results show that, by losing some acquiring performance, the algorithm can reduce the calculation resources to complete acquisition in one sequence period. And the reduced calculation resources increase with the sequence period.