The recursive least squares identification algorithm (RLS) for large scale multivariable systems requires a large amount of calculations, therefore, the RLS algorithm is difficult to implement on a computer. The comp...The recursive least squares identification algorithm (RLS) for large scale multivariable systems requires a large amount of calculations, therefore, the RLS algorithm is difficult to implement on a computer. The computational load of estimation algorithms can be reduced using the hierarchical least squares identification algorithm (HLS) for large scale multivariable systems. The convergence analysis using the Martingale Convergence Theorem indicates that the parameter estimation error (PEE) given by the HLS algorithm is uniformly bounded without a persistent excitation signal and that the PEE consistently converges to zero for the persistent excitation condition. The HLS algorithm has a much lower computational load than the RLS algorithm.展开更多
Gradient-based iterative algorithm is suggested for solving a coupled complex conjugate and transpose matrix equations. Using the hierarchical identification principle and the real representation of a complex matrix, ...Gradient-based iterative algorithm is suggested for solving a coupled complex conjugate and transpose matrix equations. Using the hierarchical identification principle and the real representation of a complex matrix, a convergence proof is offered. The necessary and sufficient conditions for the optimal convergence factor are determined. A numerical example is offered to validate the efficacy of the suggested algorithm.展开更多
The Internet today originates from the advanced research project agency network (ARPANET) and evolves constantly during the past years. However, with incremental 'patches' on the Internet, the current network arch...The Internet today originates from the advanced research project agency network (ARPANET) and evolves constantly during the past years. However, with incremental 'patches' on the Internet, the current network architecture faces many problems which are too difficult to solve. In the light of recent interest in re-designing the Internet, the separation of core network and customer edge network, and the separation of identifier and locator receive wide consensus among scientists. We follow this direction and propose a new hierarchical network with diversified customer edge access (HNDCE) in this paper. We use switching addresses, routing addresses and identifiers to implement a true core/edge and locator/identifier split. In the core network, hierarchical switching addresses combined with tree topology makes routing replaced by switching. The Viliant load-balancing (VLB) structure is used here to compensate the topology's reliability and scalability. In the customer edge network, the split of locator and identifier makes it possible to deploy various routing schemes and new services. Hierarchical mapping system is also designed to support this new architecture.展开更多
基金Supported by the National Natural Science Foundationof China (Nos.6 0 0 740 2 9and 6 9934 0 10 ) and theFoundation of Inform ation School of TsinghuaU niversity
文摘The recursive least squares identification algorithm (RLS) for large scale multivariable systems requires a large amount of calculations, therefore, the RLS algorithm is difficult to implement on a computer. The computational load of estimation algorithms can be reduced using the hierarchical least squares identification algorithm (HLS) for large scale multivariable systems. The convergence analysis using the Martingale Convergence Theorem indicates that the parameter estimation error (PEE) given by the HLS algorithm is uniformly bounded without a persistent excitation signal and that the PEE consistently converges to zero for the persistent excitation condition. The HLS algorithm has a much lower computational load than the RLS algorithm.
文摘Gradient-based iterative algorithm is suggested for solving a coupled complex conjugate and transpose matrix equations. Using the hierarchical identification principle and the real representation of a complex matrix, a convergence proof is offered. The necessary and sufficient conditions for the optimal convergence factor are determined. A numerical example is offered to validate the efficacy of the suggested algorithm.
基金supported by the National Basic Research Program of China (2007CB310701, 2011CB302900)the National Key Science and Technology Projects (2010ZX03004-002)
文摘The Internet today originates from the advanced research project agency network (ARPANET) and evolves constantly during the past years. However, with incremental 'patches' on the Internet, the current network architecture faces many problems which are too difficult to solve. In the light of recent interest in re-designing the Internet, the separation of core network and customer edge network, and the separation of identifier and locator receive wide consensus among scientists. We follow this direction and propose a new hierarchical network with diversified customer edge access (HNDCE) in this paper. We use switching addresses, routing addresses and identifiers to implement a true core/edge and locator/identifier split. In the core network, hierarchical switching addresses combined with tree topology makes routing replaced by switching. The Viliant load-balancing (VLB) structure is used here to compensate the topology's reliability and scalability. In the customer edge network, the split of locator and identifier makes it possible to deploy various routing schemes and new services. Hierarchical mapping system is also designed to support this new architecture.