|
|
|
| A Fast Online SVM Algorithm for Multi-user Detection |
| Wang Zi-lei; Xi Hong-sheng; Zhao Yu; Sheng Yan-min |
| School of Information Science and Technology, Unv. of Science & Technology of China, Hefei 230027, China |
|
|
|
|
Abstract The runtime of conventional SVM-MUD is too long to satisfy the requirement of real-time application. A fast algorithm based on online training of SVM (FOSVC) for multiuser detection is proposed in the paper. The algorithm distinguishes new added samples and constructs the current training data set using KKT condition in order to reduce the size of training samples. As a result, the training speed is effectively increased. Simulation results illustrate that the algorithm has a faster training speed and a smaller number of support vectors preserving the same quality of separating hyperplane. The performance of the FOSVC detectors is pretty much the same thing as that of SVM detectors, and much better than that of MMSE detectors.
|
|
Received: 22 November 2004
|
|
|
|
|
|
|
|
|
2006, Vol. 28 
