An efficient fault diagnosis technique for home unified service system

Home services have been emerging as a new profit for operators and a variety of home services are now being deployed. Service failures in millions of home networks impose new challenges to service operators due to the lack of operations and maintenances. In this paper, we investigate the use of a classic fault diagnosis technique, named codebook technique, for the fault diagnosis in home unified service system (HUSS) which uses a platform based technique to provide a variety of home services. To reduce the management cost and obtain the desired level of robustness, we propose two efficient minimum codebook algorithms to find the minimum set of symptoms. We evaluate the technique through a test bed of HUSS and intensive simulations. The evaluated results show that it can localize the service failures effectively and efficiently.