Browsing by Author "CHIU, JH"

Now showing 1 - 3 of 3

A SIMULATION STUDY ON THE SERVICE STRATEGIES FOR PACKET VOICE COMMUNICATION-NETWORKS
(1991-10-07)
CHANG, CJ
;
WONG, JW
;
CHIU, JH
;
電信工程研究所
;
電信研究中心
Two service strategies for packet voice communication networks are studied by simulation in this paper. The first one is the first-packet-first-priority (FPFP) service strategy. This strategy gives the packet firstly generated in a talkspurt the first priority. The second one is the maximum-delay-first-serve (MDFS) service strategy. This strategy gives the more delayed voice packet the higher priority. The FPFP and the MDFS strategies are proposed in order to match with the null timing information (NTI) and the complete timing information (CTI) reassembly schemes adopted at a packet voice receiver in Barberis and Pazzaglia's "Analysis and optimal design of a packet voice receiver", respectively, to improve the network performance. These two strategies are also compared with the simple first-in-first-out (FIFO) service strategy. The results show that the MDFS strategy has the smallest delay variation and loss rate of voice packet, the FIFO strategy has the middle values, while the FPFP strategy has the maximum.
A SIMULATION STUDY ON THE SERVICE STRATEGIES FOR PACKET VOICE COMMUNICATION-NETWORKS
(1991-10-07)
CHANG, CJ
;
WONG, JW
;
CHIU, JH
;
電信研究中心
;
Center for Telecommunications Research
Two service strategies for packet voice communication networks are studied by simulation in this paper. The first one is the first-packet-first-priority (FPFP) service strategy. This strategy gives the packet firstly generated in a talkspurt the first priority. The second one is the maximum-delay-first-serve (MDFS) service strategy. This strategy gives the more delayed voice packet the higher priority. The FPFP and the MDFS strategies are proposed in order to match with the null timing information (NTI) and the complete timing information (CTI) reassembly schemes adopted at a packet voice receiver in Barberis and Pazzaglia's "Analysis and optimal design of a packet voice receiver", respectively, to improve the network performance. These two strategies are also compared with the simple first-in-first-out (FIFO) service strategy. The results show that the MDFS strategy has the smallest delay variation and loss rate of voice packet, the FIFO strategy has the middle values, while the FPFP strategy has the maximum.
THE DELAY ANALYSIS IN A NOISY GENERAL CUT-THROUGH SWITCHING NETWORK
(1993-10-01)
CHANG, CJ
;
CHIU, JH
;
LIN, SJ
;
電信研究中心
;
Center for Telecommunications Research
The transfer delay of data packet in a noisy n-node tandem network with general cut-through switching is analyzed in this paper. By assuming each node as an M/G/1 and considering error control done only at the store-and-forward node and done at every node, respectively, the transfer delays of a data packet are determined. Both delays are illustrated and compared in numerical examples. Also, proof of the equality of the transfer delays for the two error control strategies in error free network are conducted to help verifying the validity of the analysis.