The artificial lateral line (ALL) is a bioinspired flow sensing system for underwater robots, comprising of distributed flow sensors. The ALL has been successfully applied to detect the undulatory flow fields generated by body undulation and tail-flapping of bioinspired robotic fish. However, its feasibility and performance in sensing the undulatory flow fields produced by human leg kicks during swimming have not been systematically tested and studied. This paper presents a novel sensing framework to investigate the undulatory flow field generated by swimmer’s leg kicks, leveraging bioinspired ALL sensing. To evaluate the feasibility of using the ALL system for sensing the undulatory flow fields generated by swimmer leg kicks, this paper designs an experimental platform integrating an ALL system and a lab-fabricated human leg model. To enhance the accuracy of flow sensing, this paper proposes a feature extraction method that dynamically fuses time-domain and time-frequency characteristics. Specifically, time-domain features are extracted using one-dimensional convolutional neural networks and bidirectional long short-term memory networks (1DCNN-BiLSTM), while time-frequency features are extracted using short-term Fourier transform and two-dimensional convolutional neural networks (STFT-2DCNN). These features are then dynamically fused based on attention mechanisms to achieve accurate sensing of the undulatory flow field. Furthermore, extensive experiments are conducted to test various scenarios inspired by human swimming, such as leg kick pattern recognition and kicking leg localization, achieving satisfactory results.

Entity search is a new search pattern that return related entities to users rather than amounts of web pages containing mass and messy information. It is also a challenging research topic because it is difficult to understand the meaning of users' input and identify the entities from the messy web pages. In this paper, we propose an entity search pattern based on online encyclopedias and define it as MMK search(Multi-modifier Search), which means the input text by people only includes one kernel concept and multiple modifiers. We propose a solution framework to solve this kind of search, and propose a method to identify expected entities based on well-utilized online encyclopedias. To evaluate the methods, we create an experimental data set and a baseline under the help of participants, the results verified the effectiveness of our methods.

With the booming development of social media in recent years, researchers have begun to pay more attention to extracting personal profiles from information. Keyword extraction plays an important role in extracting personal profiles. However, most of the previous studies are only valid for ordinary text, but not ideal for social media short text. In this paper, we propose an improved method for keyword extraction based on Word2vec and Textrank to solve the unique problem of social media short text. Our approach uses the Word2vec to capture the semantic features between words in selected text, and meanwhile naturally fuses the word frequency, semantic relation and directional relation into Textrank to extract keywords. We conduct the experiments on the three datasets. The experimental results show the superior performance of our method in keyword extraction.

With the development of Internet, various kinds of new applications appear constantly. They all have high requirements to the time delay, throughput, especially strong real-time applications such as mobile monitoring, video calls. This is a new challenge to the existing congestion control method. In order to solve this problem, we propose novel adaptive queue management intelligent algorithm in this paper. New active queue management algorithm adopts a new formula to calculate the discard packet rate. The discard packet rate can be calculated according to the changes of average queue and the nonlinear function. This new algorithm named ASRED (Adaptive Sigmoid RED) is based on the framework of RED (Random Early Detection). ASRED uses a new function to calculate the discarding probability. In addition, the adaptive adjustment of maxp mechanism is added into the algorithm.

One of the key problems that should be solved is the collision between tags which lowers the efficiency of the RFID system. The existed popular anti-collision algorithms are ALOHA-type algorithms and QT. But these methods show good performance when the number of tags to read is small and not dynamic. However, when the number of tags to read is large and dynamic, the efficiency of recognition is very low. A novel approach to mapped correlation of ID for RFID anti-collision has been proposed to solve the problem in this paper. This method can increase the association between tags so that tags can send their own ID under certain trigger conditions, by mapped correlation of ID, querying on multi-tree becomes more efficient. In the case of not too big number of tags, by replacing the actual ID with the temporary ID, the method can greatly reduce the number of times that the reader reads and writes to tag's ID. In the case of dynamic ALOHA-type applications, the reader can determine the locations of the empty slots according to the position of the binary pulse, so it can avoid the decrease in efficiency which is caused by reading empty slots when reading slots. Experiments have shown this method can greatly improve the recognition efficiency of the system.