Three Professors of IT Faculty Obtained 4 Innovation Patents

The professors of Faculty of Information Technology, Professor Liang Yong,Associate Professor Zhao Qing Lin, Assistant ProfessorFeng Li,and their research teams obtained 4 Innovation Patents certified by Australian Government, and are listed as follows


1.Semi-Supervised Learning Framework based on Cox and AFT Models with L1/2 Regularization for Patient’s Survival Prediction, LIANG, Yong; CHAI, Hua and LIU, Xiao-Ying, 9/25/2015.

We designed a novel semi-supervised learning method based on the Cox and AFT models to accurately predict the treatment risk and the survival time of the patients. Moreover, we adopted the efficient L1/2 regularization approach in the semi-supervised learning method to select the relevant genes, which are significantly associated with the disease. The advantages of our proposed semi-supervised learning method include: 1) significantly increase the available training samples from censored data; 2) high capability for identifying the survival risk classes of patient in Cox model; 3) high predictive accuracy for patients’ survival time in AFT model; 4) strong capability of the relevant biomarker selection. Consequently, our proposed semi-supervised learning model is one more appropriate tool for survival analysis in clinical cancer research.



2.A Single Image Super-Resolution Method Using Transform-Invariant Directional Total Variation with S1/2+L1/2-norm, LIANG, Yong; XU, Zong Ben; XIA, Liang-Yong and LIU, Xiao-Ying, 9/17/2015.


We proposed a novel TI-DTV model with Schattenp=1/2 (S1/2-norm) and L1/2-norm penalties (TI-DTV+S1/2+L1/2), in which S1/2 -norm and L1/2-norm are used to induce the lower rank and more sparse components for image super-resolution. Numerous experiments show the proposed method can make the horizontal and vertical straight edges sharpness, reduce the jagged artifacts along the diagonal line and arcs, and recover more information to high quality super-resolution using single low resolution image.



3.A Novel Energy-Efficient Scheme for Coding-Aware Routing, ZHAO, Qinglin; KAI, Caihong and ZHENG, Hanxu, 9/24/2015.

Wireless sensor networks have wide applications such as water quality monitoring and natural disaster prevention. Energy-efficient designs are essential in order to maximize the lifetime of wireless sensors. In this patent, we propose a network-coding-aware routing scheme to efficiently reduce the energy cost in wireless sensor networks. The basic ideas are as follows: when multiple unicast traffic-flows are delivered, by carefully assigning traffic flows among different paths and exploring the chances of network coding, we can reduce the times of packet relaying and transmission, thereby decreasing energy consumption. Simulation results show that the proposed scheme has great potential of reducing 11% of energy consumption compared to traditional non- network-coding-aware routing schemes. This patent is very useful in extending the lifetime of wireless sensor networks.



4.A Novel Method for Optimizing System Parameters in Delayed Channel Access Protocol, Feng Li.2015/9/1


The packet aggregation technology is a salient feature of IEEE standards, 802.11n and 802.11ac. With this technology, multiple packets are packaged into a super-frame for one transmission and therefore the channel utilization is improved. One basic question is: how long a node should be waited for so as to aggregate more packets for one transmission. In this patent, we propose a theoretical model to calculate the optimal waiting delay so as to improve the efficiency of the packet aggregation. Extensive ns2 simulations verify that our model is very accurate and our method can maximize the system throughput. This patent is very useful in improving and optimizing the packet aggregation technology.